This research evaluates efficiency levels of individual paratransit systems in Canada with the specific objective of identifying the most efficient agencies and the sources of their efficiency. Through identification of the most efficient systems along with the influencing factors, new service policies and management and operational strategies might be developed for improved resource utilization and quality of services. The research applies the data envelopment analysis methodology, which is a mathematical programming technique for determining the efficiency of individual systems as compared with their peers in multiple performance measures. Annual operating data from 2001 to 2003 as reported by the Canadian Urban Transit Association are used in this analysis. A bootstrap regression analysis is performed to identify the possible relationship between the efficiency of a paratransit system and measurable operating or managerial factors that affect the performance of paratransit systems. The regression analysis allows for the calculation of confidence intervals and bias for the efficiency scores.