AbstractThere has been a growing interest in developing and implementing animal welfare assessment schemes on farm, in zoos, in experimental situations and even in the wild. Scientists, inspection bodies and politicians are also starting to consider seriously the use of animal welfare outcome-based measures (OBMs) such as behaviour or physical condition as a progression from resource-based measures (RBMs) for such schemes. Measures based directly on the animals can provide good indicators under a variety of conditions, as welfare is a characteristic of the individual animal, not just of the system in which animals are kept. Many modern farms, zoos, aquaria and other facilities are keen to identify welfare-monitoring tools that can be readily applied and allow for rapid assessment to provide timely feedback for management decisions. Improvements in welfare can be achieved through the combination of: (i) measurement; (ii) analysis of risk and environmental factors; (iii) provision of information resulting from the assessment; and (iv) promotion of positive change by supporting management decisions. In principle, individual measures can also be combined to give aggregate scores that can be presented to the producer, the animal keeper or the consumer. This requires the attribution of weighted values to the measures used to assess the impact of each measure with respect to animal welfare. Four questions arise about any approach that assesses the animals themselves: is it practical, is it valid (providing 'real' information about welfare), is it repeatable and is it robust (not influenced by weather, etc.)? Yet animal experience cannot be reduced to a mechanistic assessment: animals are variable, living and sentient beings and this must be realistically addressed in practical assessment systems. A multi-pronged strategy involving various RBMs and OBMs is most likely to provide the capacity for comprehensive welfare monitoring. Such a strategy may include regularly updated guidelines for species care, accreditation standards, longitudinal, multi-disciplinary and multi-institutional studies, and assessment tools for continuous welfare monitoring. Two case studies are discussed in detail: welfare assessments of broiler chickens on farms and of clouded leopards in zoos. A laboratory animal application is then described related to the implementation of humane end points for research.