Standard measures used in animal welfare assessments include animal-based measures obtained by observing animals, resource-based measures obtained by observing facilities, and management-based measures obtained by interviewing farmers and checking records. Animal welfare assessments are composed of a variety of measures and can be very different from each other. The objectives of the research presented in this thesis were to investigate inter-observer reliability of three swine welfare assessment programs (ACA™, PQA Plus®, and Welfare Quality®), and to determine the concordance of rankings across the 3 assessments. All assessments and all types of measures were found to be highly reliable. Moderate concordance was found for the rankings of farms across all three assessments. The lowest ranked farms were correlated, indicating that all assessments identified the farms with the lowest levels of animal welfare. Results of this study can be used to identify the best measures and revise current on-farm animal welfare assessments.