The efficacy of a nursing home Safe Resident Handling Program (SRHP) to reduce the ergonomic exposures of nursing assistants was evaluated. The healthcare version of the PATH method was used by 12 observers to examine postures, manual handling, and resident handling pre-intervention and at three months, 12 months, 24 months, and 36 months post-intervention. There were marked downward trends in proportion of work time spent repositioning and transferring residents, and increased use of handling equipment in transferring (Cochran-Armitage tests: all p-values < 0.001). While resident handling post-intervention, nursing aides were more likely to be in neutral trunk postures, walking rather than standing still, working with both arms below 60 degrees, and less likely to lift loads greater than 50 pounds. Lateral transfer devices were infrequently observed in use for repositioning; additional training on the use of this equipment is recommended to increase the potential benefits of the intervention program. A biomechanical index was developed that combined the compressive forces on the spine resulting from the observed postures and manual handling, in order to obtain a comprehensive analysis of the physical workload of nurses and nursing assistants in long-term care facilities. Informed by a prior biomechanical model that incorporated workers' self-reported frequencies of postures and manual handling, observational data of ergonomic job features was used. The University of Michigan's Three-Dimensional Static Strength Prediction Program (3DSSPP) was used to calculate compressive forces on the lumbar spine resulting from 17 combinations of trunk, arm, and leg postures and manual handling activities. Each force estimate was then used as a weight for the observed frequency of that combination of PATH variables by job group, and the contributions were summed to obtain total physical loads. These total loads were computed for the four observational surveys from before to three years after the ergonomics intervention. Over the follow-up period the physical workload index (PWI) decreased both for nursing assistants (-24.2%) and for nurses (-2.5%). The index for nursing assistants was much higher during resident handling than other tasks. By the end of follow-up, the index for nursing assistants while resident handling decreased by 40.9% of the baseline value. Differences in the efficacy of the SRHP in five of the nursing homes in the sample were examined. Two outcome measures were considered: changes in equipment use while resident handling and changes in the PWI for nursing assistants over a two-year period following SRHP implementation. Questionnaires, administrative data, employee satisfaction surveys, and staff exit interviews following the collection of ergonomic observations were examined for explanatory factors of between-center differences in outcomes. Of the explanatory factors, significant correlations related to the outcome measures were the percentages of agency staff used to fill shifts, work shifts involving obstacles to getting work done, 'never' feeling time pressure, adequacy of supplies and equipment, 'poor' ratings for quality of teamwork and staff-to-staff communication, and observed understaffed shifts. The facility with the most positive outcome measures was associated with many positive changes in explanatory factors and the facility with the least positive outcome measures experienced negative changes in the same explanatory factors. These explanatory factors might also inform future analysis of the outcome measures on individuals.