Search scientific articles, books, and other publications related to improving or safeguarding the welfare of animals used in research and testing.

Background: Efficient head fixation is critical in primate behavioral experiments for accurate eye tracking and neural studies. Existing methods risk infection and bone necrosis due to their invasive nature. New method: We propose a custom-designed, 3D-printed non-invasive head immobilization system (NHIS) for macaques based on precise CT/MRI scans and integrated with our specialized software, FLoRIN, and Blender 3D. This NHIS is tailored for straightforward manufacturing and personalization and accommodates the necessary experimental equipment. Results: This innovative NHIS ensures reliable head stabilization while significantly reducing invasive procedure risks. The design simplifies the setup process, curtails infection risk, mitigates bone damage, and increases comfort for subjects. Comparison with existing method(s) Our NHIS surpasses previous invasive and non-invasive methods by removing the need for surgical implantation, extensive training, and discomfort. It is designed for ease of adjustment and fitting without sedation, enhancing adaptability and animal welfare. Conclusions: The NHIS signifies a significant advancement in non-invasive primate head fixation, providing a customizable, less invasive, and more reliable solution for research. Its potential for widespread adoption in diverse research scenarios promises to improve the subject well-being and the quality of data collected.

A multicenter study was undertaken to generate new weight-for-age growth standards to monitor the growth of infant rhesus and pigtail macaques in research breeding colonies. Previously, single institutions have developed reference growth curves, under the assumption of linear growth, for infants raised in certain settings with a limited scope of benefit to outside institutions. Weight and health records from seven research institutions across the USA were used to build models of body weight by age. Linear and Box-Cox Power Exponential (BCPE) distribution models, which have been adopted from the World Health Organization’s methods, were compared to find the best fit of the models. Resultant weight percentiles and growth velocity charts were provided from the best-fit model ranked by generalized Akaike’s information criterion. Multiparameter models with the BCPE distribution fit the data better than linear models with normal distributions. The assumption that infant macaque growth is linear was challenged by our findings; growth rates appear to change over the first year of life for infant macaques. Growth standards were integrated into centile charts and a computer application in MS Excel to provide user-friendly weight-for-age percentile charts to access and analyze macaque growth data. The new growth standards provide a unified reference that best represents normal physiological growth for all infant rhesus and pigtail macaques from birth through the first year of life. These standards offer guidance on expected growth trajectories and serve as a benchmark for assessing the healthy development of macaque infants across research breeding colonies.

Given that there were approximately 4 million cats in shelters across 20 countries in 2024, there is a critical need to have accurate, reliable, non-invasive tools to evaluate shelter cat stress and welfare. This study was conducted to validate the newly developed Psychological Quality of Life (PQOL) assessment tool to evaluate welfare in cats newly admitted to an animal shelter. The PQOL is an 11-question assessment of a cat’s behavior, attitude, and activity. The PQOL was compared to the Cat Stress Score (CSS), a widely used non-invasive assessment tool used to determine stress levels in shelter cats. Fifty cats from a local shelter were assessed with both the PQOL and CSS by two independent observers at two time points on their second, fourth, and sixth day in the shelter. A Spearman rank correlation evaluated the relationship between the overall PQOL and CSS scores and between the overall PQOL and select individual PQOL questions. The inter-rater reliability (IRR) and intra-rater reliability (IaRR) for the overall PQOL, CSS and individual PQOL questions were measured with a quadratic weighted Cohen’s kappa. There was a strong, positive correlation (0.73–0.79) between the CSS and overall PQOL for all three assessment days. The kappa for the CSS and overall PQOL had substantial agreement (0.58–0.75) between reviewers for all three assessment days at both time points indicating good reliability. Individual PQOL questions about interest in play, engagement with others, exploration, and showing positive emotions had a strong correlation with high overall PQOL for both time points across all three assessment days. The PQOL question about showing fear or anxiety had a strong correlation with a low overall PQOL score. The overall PQOL score was shown to be a valid and reliable metric to monitor stress and welfare in shelter cats.

Training is an essential part of meeting the needs of elephants in human care. In particular, training in a protected contact environment inherently offers more control over the environment and the choice to perform requested behaviors, likely improving welfare. Here we present a follow-up assessment of caregiver and elephant behavior nearly 20 years after the initial transition from free to protected contact to explore any longitudinal changes and the potential effects of moving to a newly constructed exhibit. As with the previous study examining the initial transition to protected contact, the present study reports high levels of compliance with caregiver requests, even after nearly two decades and through a major transition in the environment. We propose that implementing longitudinal studies to periodically evaluate training programs and develop best practices for meeting the needs of elephants in human care are an important contributor to elephant welfare.

Background: Spared nerve injury (SNI) is a widely used rodent model in the study of peripheral neuropathic pain. The injury, that involves ligation and resection of the common peroneal and tibial branches of the sciatic nerve, results in a robust and prolonged nociceptive response. The model is characterised by measurable mechanical and thermal hypersensitivity in the innervation area of the spared sural nerve of the lateral hind paw, that persists for several months. Since its description, the model has been used to further the understanding of a diverse range of mechanisms contributing to the development and maintenance of neuropathic pain. However, due to damage to the biceps femoris muscle during surgery, experimental results may not reflect the mechanisms of neuropathic pain alone. New method: Here we describe an alternative protocol for SNI in the mouse, which involves manipulation of the sciatic nerve via a medial thigh incision. Results: Mice which have undergone the medial thigh SNI procedure exhibit molecular and behavioural changes as observed in the lateral thigh SNI procedure, including glial activation and persistent mechanical, cold and heat hypersensitivity of the affected limb. Comparison with existing methods: The medial thigh approach promotes ease of the surgical procedure and benefits animal welfare through improved wound healing and preservation of the biceps femoris muscle, when compared to the lateral thigh SNI model of peripheral neuropathic pain. Conclusions: We present an alternative protocol for the SNI procedure which offers several advantages when compared to the previously described lateral thigh procedure.

Background The blood flow changes due to various physiological or pathological causes represent the response of the CNS for the maintenance of metabolic and functional requirements of the brain and related structures. Visualization of blood dynamics can enable the monitoring of crucial responses of neuronal tissue and the vasculature to multiple stimuli or insults. Here, we propose an inexpensive and feasible method that can be easily implemented to observe cerebral blood flow dynamics and for heart rate estimation in mice. New Methods We developed a custom, MATLAB-based image processing software running at the pixel level to estimate regional cerebral blood flow (CBF) changes from acquired RGB images, also capable of modeling simultaneously operating Laser Doppler Flowmetry (LDF) readouts and heart rate estimation in mice. Results Proposed method clearly enabled the visualization of vascular changes associated with cortical spreading depolarizations (CSD) induced in mice. The analysis of RGB intensity variations within pre-definable ROIs allowed the quantitative comparison of blood flow changes. The employment of an artificial neural network (ANN) yielded to successful approximation to simultaneously acquired LDF readings. The heart rates of the mice were also calculated within close proximity to those of obtained by ECG. Comparison with existing methods Our method can help reduce the expenses of blood flow monitoring, where previously described methods are not readily accessible, with additional benefit of heart rate estimation in mice. Conclusion This study suggests a cost-effective and efficient approach for monitoring vascular responses where applicable in relevant research areas.

The use of Vascular Access ButtonsTM (VABs) is gaining traction in animal research as a less invasive alternative to traditional venipuncture methods and a superior option compared with existing alternatives, such as vascular access ports (VAPs). By enabling repeated blood collection and intravenous drug administration with reduced stress and discomfort for the animals, VABs contribute to improved data quality and enhanced animal welfare in animal studies. While increasingly used in rodents, their application in non-rodent species remains underexplored. This study evaluated the feasibility and reliability of the VAB system in two Göttingen Minipigs, comparing its performance with the conventional vena cava cranialis puncture. Pharmacokinetic assessments and clinical pathology analyses revealed consistent results across both techniques, demonstrating the VAB system’s ability to generate reproducible, high-quality data. Additionally, its durability and ease of use highlight its potential as a practical and ethical alternative to both traditional venipunctures and VAPs in pharmacokinetic and long-term studies in minipigs. These findings support the integration of the VAB system in toxicological and pharmacokinetic research, particularly in studies requiring repeated blood collection or chronic intravenous dosing.

In laboratory settings, human–rat interactions are necessary and frequent, yet the impact of different handlers on the behaviour of the animal is underexplored. This study examined how human behavioural responses - confidence, experience, and physiological arousal - relate to affective indicators in rats, specifically ultrasonic vocalisations (USVs) and behavioural responses. 50-kHz USVs, associated with positive affect, were significantly higher post-handling compared to during handling, but no aversive 22-kHz calls were observed. Correlation analyses found that rat behavioural distress scores were positively associated with handler behavioural distress scores and negatively associated with handler’s confidence. Heart rate increased during handling, particularly among moderately experienced handlers (1–5 years), but was not associated with rats’ behavioural distress measures. Experience was positively correlated with confidence but not directly associated to rat behaviour or handlers heart rate. These findings suggest that observable human behaviour, rather than physiological arousal or years of experience, can potentially influence rats’ emotional responses with the potential implications for animal welfare and scientific outcomes.

In laboratory settings, human–rat interactions are necessary and frequent, yet the impact of different handlers on the behaviour of the animal is underexplored. This study examined how human behavioural responses - confidence, experience, and physiological arousal - relate to affective indicators in rats, specifically ultrasonic vocalisations (USVs) and behavioural responses. 50-kHz USVs, associated with positive affect, were significantly higher post-handling compared to during handling, but no aversive 22-kHz calls were observed. Correlation analyses found that rat behavioural distress scores were positively associated with handler behavioural distress scores and negatively associated with handler’s confidence. Heart rate increased during handling, particularly among moderately experienced handlers (1–5 years), but was not associated with rats’ behavioural distress measures. Experience was positively correlated with confidence but not directly associated to rat behaviour or handlers heart rate. These findings suggest that observable human behaviour, rather than physiological arousal or years of experience, can potentially influence rats’ emotional responses with the potential implications for animal welfare and scientific outcomes.

Introduction Selecting an appropriate analgesic for cats can be challenging due to potential unwanted side effects, short duration of action, or unsuitable presentation for home administration. This study aimed to evaluate the analgesic efficacy and safety of a fixed-dose oral combination of dipyrone and tramadol (Sindolor Cats® tablets) for postoperative pain control in cats undergoing ovariohysterectomy. Methods Thirty-six healthy female cats were randomly assigned to three groups (n = 12): GTD (dipyrone 12.5 mg/kg + tramadol 2 mg/kg), GT (tramadol 2 mg/kg), and GD (dipyrone 12.5 mg/kg). Treatments were administered orally every 12 h for 5 days. Anesthesia was induced with propofol and maintained with isoflurane; all cats also received sacrococcygeal epidural lidocaine under ultrasound guidance. Pain was assessed using the CMPS-Feline and the Feline Grimace Scale (FGS). Physiological parameters (heart rate, respiratory rate) and serum cortisol were measured at baseline and at 2, 4, 6, and 24 h postoperatively. Rescue analgesia with morphine was administered when pain scores exceeded threshold values. Results The GTD group required less rescue analgesia (1/12) compared to GT (4/12) and GD (5/12). Although not statistically significant, pain scores were consistently lower in GTD. Cortisol concentrations were significantly reduced in GTD at 4 h compared to GT and GD. Side effects such as sialorrhea were markedly lower in GTD (2/12) versus GT (9/12) and GD (12/12). No signs of cardiovascular or respiratory depression were observed in any group. Discussion The combination of dipyrone and tramadol in a single oral formulation provided superior analgesia and fewer adverse effects than either drug alone. Cats in the combination group required less rescue medication, had significantly lower cortisol values, and experienced fewer side effects.

Introduction With the growing number of veterinary centers performing pacemaker implantation in small animals, a better understanding of anesthetic management of these procedures has become essential. Most available studies focus primarily on surgical techniques, with limited discussion of anesthetic safety, efficacy, and intraoperative complications. Materials and methods Eighteen anesthetic procedures for transvenous pacemaker implantation were retrospectively reviewed in dogs treated at a veterinary referral center in São Paulo, Brazil, between 2024 and 2025. All dogs underwent clinical and laboratory evaluation, electrocardiography and/or Holter monitoring, and echocardiography. Continuous intraoperative monitoring included heart rate, electrocardiography, respiratory rate, end-tidal CO2, oxygen saturation, temperature, and invasive or non-invasive arterial blood pressure. Collected data included age, breed, sex, anesthetic protocol, clinical conditions, intraoperative events, and outcomes. Continuous variables were expressed as mean ± SD or median (range) according to data distribution assessed by the Shapiro–Wilk test. Group comparisons were performed using Student’s t-test or Mann–Whitney U test, and categorical variables were analyzed using chi-square or Fisher’s exact test (p < 0.05). Results Sixteen dogs (8 males and 8 females) underwent 18 anesthetic procedures, including two reinterventions. The main indication for pacemaker implantation was third-degree atrioventricular block (75%). Anesthetic protocols were individualized according to patient comorbidities. The most frequent anesthetic complication was hypotension (27.7%), successfully managed with fluid therapy and vasoactive support. Electrode migration occurred in 11% of procedures and required reintervention. No perioperative deaths were recorded. Discussion and conclusion Transvenous pacemaker implantation in dogs was successfully performed under individualized anesthetic management despite heterogeneous clinical conditions and absence of a standardized protocol. Hypotension was the most common anesthetic complication but was effectively treated. The routine use of temporary external pacing provided an essential safety margin for maintaining perioperative hemodynamic stability. No perioperative deaths were recorded. These findings support the feasibility and safety of tailored anesthetic strategies for canine pacemaker implantation, although prospective studies with larger populations are warranted to define optimal standardized protocols.

Wearable and implantable physiological monitors are designed for minimally invasive health assessment in companion animals. We aimed to establish proof-of-concept for the use of PetPace collar monitors and VetChip implantable monitoring devices in pigs. PetPace monitors were used to measure pulse rate, respiratory rate and heart rate variability in individually housed pigs. Group housed pigs were fitted with PetPace monitors to assess their feasibility for use in the presence of conspecifics, in addition to implantation with VetChip subcutaneous monitoring devices. PetPace monitors caused skin abrasion after 10 days of continual wear, but simple modifications allowed collars to be worn safely in individually housed pigs. Collars could not be worn unsupervised by group housed pigs due to damage by conspecifics. PetPace pulse and respiratory rate readouts showed a strong correlation with manual auscultations, and PetPace heart rate variability showed a very strong negative correlation with salivary corticosterone levels. VetChip temperature and pulse rate readouts were accurate in healthy pigs. However, the suitability of dorsal mid-scapula VetChip implantation is questionable due to wound breakdown in two out of three pigs.

Assessing laboratory animals’ welfare – their current and/or past subjective affective states – is essential for ethical and regulatory reasons (and central to biomedical research into, for example, pain, nausea or anxiety). But this is challenging; and in the quest for quantification (and perhaps simplicity), it can be tempting to overlook construct validity. Nevertheless, that our indicators have good construct validity – that is, they accurately reflect the construct or concept of interest – is essential. This is true whether we are interested in short-term emotions like fear, longer-term mood-like states such as malaise, or markers of cumulative stress over a project or even a lifespan. Without it, welfare assessments risk being incorrect: inaccurate and unhelpful for the animals they aim to evaluate and assist. Here (summarising text from a forthcoming edited book), I introduce five validatory tests, as well as highlighting the importance of considering indicators’ responsiveness/sensitivity and selectivity/specificity. I also outline how these principles could help improve the construct validation of both humane endpoints and retrospective severity assessments. Careful construct validation can never fully solve the ‘Other Minds’ problem: that animals’ subjective experiences are private (such that we can never measure them, only infer them). However, done well, construct validation would add additional logical rigour to laboratory animal welfare assessment, increase its accuracy, and make benchmarking (e.g. severity classification) more transparent.

Despite the substantial growth of cephalopod research over the past 15 years, effective pain management for these animals remains largely unstudied, and no US federal regulations currently address cephalopod analgesia. This 2-phase study evaluated the feasibility and efficacy of lidocaine, bupivacaine, and extended-release bupivacaine injections for producing analgesia in juvenile Octopus bimaculoides in an established model of arm regeneration. Phase I served as a pilot to assess adverse effects, practicality, and histopathologic consequences of drug administration. Forty-nine animals were randomized to the saline control or treatment groups. Phase II assessed lidocaine, bupivacaine, and saline in an additional 34 animals, using behavioral metrics at multiple time points, including a posture and gait analysis during a walking assessment and mechanical allodynia evaluation via the von Frey filament test. Health indicators, such as weight change and respiratory rate, were also monitored. Finally, wound healing and arm regeneration were followed with photo documentation. Although all animals displayed statistically significant changes in arm posture and gait immediately postsurgery, these changes invariably resolved within 72 hours irrespective of treatment group. Moreover, health and regenerative outcomes did not differ significantly among the treatment groups. The failure of local anesthetic efficacy highlights the need for alternative analgesic regimes for octopus research.

Providing nesting material is one of the main forms of environmental enrichment for laboratory mice. Nest building is an innate behavior, widely observed in wild mice, which construct complex, dome-shaped, multi-layered structures. Although various materials can be used in microisolator cage systems, this behavior remains underexplored under laboratory conditions. Evidence indicates that laboratory mice can build nests comparable to those of their wild counterparts, reinforcing the relevance of this enrichment for animal welfare. The aim of this study was to evaluate nest-building behavior in mice, considering biological and environmental variables, including strain, sex, age, nest location within the cage, seasonal variation, ambient temperature, and the use of cardboard tunnels. A total of 620 observations were analyzed from 106 cages, each containing 2–5 mice of both sexes, distributed across three strains (B6J, Nude, and BTBR). Nests were assessed weekly over the course of one year using a scoring system adapted from previous studies, in which higher scores indicate more elaborate, dome-shaped structures, whereas lower scores correspond to incomplete, flat, or less cohesive constructions. The use of cardboard tunnels provided as enrichment was also recorded (intact, shredded, or soiled with excreta). Other variables evaluated included age (20–109 days), and the position and side of the nest within the cage. Most mice built well-structured nests, with 39% of cages reaching the maximum score of 5. Statistical analysis showed that nest quality was significantly influenced by strain, season, and nest location within the cage. Compared to Nude mice, B6J and BTBR strains showed reduced nest-building performance. Scores were lower in spring compared to winter, while nests located at the cage edges, particularly on the right side and at the front, had slightly higher scores than those in the center. In conclusion, strain is the primary determinant of nest quality, while environmental and positional factors exert additional effects. Nude mice, during winter or colder periods, and nests located at the edges and front of the cage were associated with more complex nest constructions.

This study evaluates the procedure for collecting saliva to measure salivary cortisol levels in dogs, focusing on the potential stress induced by saliva collection in unfamiliar experimental settings. While salivary cortisol measurement is widely regarded as a non-invasive method for assessing stress in dogs, limited data exist on whether the procedure itself may cause discomfort or stress. By addressing this gap, the study aims to explore welfare-related challenges in interdisciplinary research involving non-human subjects and to contribute preliminary insights toward the development of animal-friendly, standardized methodologies. In addition, it sheds light on the organization of an interspecies study involving both human and non-human actors. The study included 19 companion dogs of various breeds, sizes, and temperaments to ensure diverse representation. Saliva samples were collected from dogs during sociological and psychological interviews, with salivary cortisol levels analyzed at the beginning and end of the session. A standardized saliva collection procedure was employed using SalivaBio Children’s Swabs (Salimetrics, State College, PA, USA). The swabs were placed under the dog’s tongue or inside the cheek pouch for approximately two minutes. The procedure was conducted in a controlled environment (a therapy room at the Institute of Psychology) to minimize external stressors. Salivary cortisol concentrations were measured using an enzyme-linked immunosorbent assay (ELISA). Behavioral observations were conducted via video recordings and analyzed by a zoopsychologist using Capra’s Ethogram (2012) as a reference framework. A total of 38 video recordings were analyzed to assess stress-related behaviors. The findings revealed that, while most dogs displayed mild to moderate stress during saliva collection, no extreme stress reactions were observed. Caregiver presence seemed to play a role in reducing visible signs of stress, with emotional support potentially mitigating dogs’ behavioral discomfort. However, these findings are based on qualitative observations and warrant further empirical investigation. Key challenges included variability in saliva production and occasional gum bleeding, which may exclude some samples from the analysis. These factors should be considered when designing future studies. Despite these challenges, salivary cortisol measurement proved to be a reliable and minimally invasive tool for assessing stress, especially when combined with behavioral observations to contextualize physiological data. This study underscores the necessity of standardized protocols that prioritize animal welfare, offering valuable insights to refine research methodologies and deepen our understanding of canine stress responses. Overall, the study proposes a protocol for saliva collection adapted for social science research involving dogs, with an innovative, structured preparatory stage designed to enhance animal comfort and minimize stress associated with the unfamiliar research environment.

Monitoring the health and well-being of research animals is essential for both ethical and scientific purposes. In songbirds, body weight is one of the main indicators of their overall condition, yet traditional weighing methods can be intrusive and stress-inducing, which could decrease their song rate. We developed an automated system for monitoring the weight of multiple birds in longitudinal neuroscience experiments, which often include birds tethered to data acquisition systems. Building on previous models that were mostly designed for weighing birds outdoors and in large housing cages, our design serves as a perch for the bird to stand on (a perch-scale) and meets the needs of neuroscience experiments by (i) minimizing cable entanglement to safely accommodate tethered birds, (ii) supporting long-term monitoring of individually-housed birds, and (iii) linking multiple devices into a unified control unit that oversees setup, calibration, and data acquisition. We deployed the system in the cages of six canaries for ten days and validated its accuracy against daily manual weighing. The precision and continuous monitoring of the perch-scale allowed observing physiological patterns such as overnight weight loss. Our system detected 22 sequences of overnight perching in five different birds, showing an average decrease of ≈4% of the bird’s body-weight overnight. We also found that daily weight estimates, derived from perch-scale data, were within the range of daily weight fluctuations (5–10%), as they deviated by less than 5% on average when compared to the manual weights. These results validate the device’s sensitivity for detecting subtle and health-related changes. By eliminating the need for manual handling of birds, this system offers a non-invasive, hands-free approach that reduces stress and improves the accuracy of health assessments. Future applications could integrate additional health metrics to provide a more comprehensive understanding of animal welfare in neurophysiology and behavioral studies.

The concept of choice is increasingly recognized as a critical component of animal welfare, particularly within zoological settings. Despite its significance, the academic literature on empirically examined methods to allow zoo animals to actively communicate choice is limited. This paper presents a novel “You Choose” procedure that allowed multiple pinniped species–including New Zealand fur seals (Arctocephalus forsteri), California sea lions (Zalophus californianus), and an Australian sea lion (Neophoca cinerea)–along with Pacific bottlenose dolphins (Tursiops truncatus) to communicate choice between concurrently-presented options through the autonomous selection of one outcome. We describe two different scenarios in which this choice procedure was implemented: (1) To select among potential rewards during a training session; or (2) to allow the animal to pick between potential enrichment items. In both cases, training the animal to utilize the “You Choose” procedure is detailed, with the purpose of providing the animal the ability to make such choices. Training zoo animals to communicate choices to keepers from an array of concurrently available stimuli should support the animals’ welfare needs, as well as introduce a framework for implementing choice-driven training and enrichment practices that can be adapted for other zoo-housed species. We propose a future methodological approach utilizing the “You Choose” procedure which allows for measurable choice manipulations, thereby suggesting future research opportunities to assess the quantitative impact of choice on welfare.

Rainbow trout (Oncorhynchus mykiss) is increasingly employed as an experimental model, yet knowledge of its behaviour in laboratory settings is limited. This study examined short-term (24 h), mid-term (1 week), and circadian changes in behavioural and spatial patterns of individually-housed trout under two experimental conditions differing in the extent of underwater current: control vs water flow. Moreover, the effects of feeding and tank cleaning were evaluated to identify potential behavioural indicators of welfare of the species in this context. Short-term effects revealed that flow rate impacted on swimming behaviour, with trout showing less species-specific behavior in the absence of water flow. Over time, trout exhibited a marked reduction of behavioral expression, interpretable as behavioural stabilization but likely reflecting suboptimal welfare. Trout behaviors were affected by the day time being less active in mornings. Event-related findings indicated an increased in active behaviors and a decrease in maintenance behaviours after both husbandry procedures, with that of cleaning being more pronounced and lasting remarkably longer, suggesting that the two reactions were fundamentally different. This study highlights the need to critically evaluate behavioural expression, and the need to provide species-specific husbandry conditions. We also provide foundational insights into rainbow trout behaviour valuable for refining research practices and improving welfare assessment of this experimental fish.

Harbour (Phoca vitulina) and grey (Halichoerus grypus) seals face numerous anthropogenic and environmental threats around the UK and Ireland. These commonly lead to seal pups becoming stranded and in need of rescue and rehabilitation. Although rehabilitation supports the recovery and welfare of stranded seals, the process can expose individuals to a range of stressors, such as handling, intensive medical treatments, assisted feeding and confinement. Effective methods for assessing seal health and welfare in response to these challenges remain limited. This study investigated whether faecal cortisol metabolites (FCMs) and behavioural indicators can effectively monitor welfare in rehabilitating harbour and grey seal pups. A total of 479 faecal samples were collected from 19 seal pups (Harbour n = 8, Grey n = 11), and 3864 h of footage was collected for 25 seal pups (Harbour n = 11, Grey n = 14), all rehabilitated at Tynemouth Seal Hospital (North Tyneside, UK). The effect of sex, time of day, feeding method, water access, body condition, and day of rehabilitation on both FCM concentrations and behavioural time budgets was investigated. Changes in FCM levels did not correspond to any of the predictor variables and highlights the complexity of interpreting FCMs in rehabilitation settings, likely reflecting individual variation, allostatic load, and limited sensitivity to short-term stressors. In contrast, seal pup behaviour was significantly affected by rehabilitation variables, including feeding method and access to water. Although assisted feeding is essential for the survival of malnourished pups, it caused decreased displays of vigilance in both species, and a decrease in tactile and locomotion behaviours with an increase in stereotypic behaviours in grey seals. This warrants enrichment plans to be incorporated into the care of these seal pups to counteract any negative impacts on their welfare. Similarly, water access for swimming should be provided as soon as possible to ensure that the sufficient development of natural locomotory and tactile behaviours is facilitated. Overall, these results could be used to optimise behavioural welfare assessments and the rehabilitation protocols currently in place for both harbour and grey seal pups, improving seal pup welfare in captive settings.