Nursing aides (18 women and six men) at one traditional and one modern geriatric ward, and warehouse workers (16 men) at two different types of warehouse departments were studied during occupational work. The vertical load during manual handling was measured using strain gauges built into wooden shoes. The warehouse workers performed four times as many lifts as the nursing aides and transferred five times as great a mass per unit time. Less than 25% of the lifts were carried out with the load evenly distributed on both feet during both the upward lift and lowering. In the warehouses the lifts were short, while the nursing aides were exposed both to lifts of longer duration and to carrying, as well as a greater frequency of unexpected, sudden and high peak load. The heart rate and oxygen uptake values recorded were relatively low, and both groups utilized approximately 25% of their maximum aerobic power. According to official occupational injury statistics, both warehouse workers and nursing aides belong to risk groups with a high frequency of reported back injuries. However, more injuries are reported by nursing aides than by warehouse workers. Greater muscle strength in male workers may partly explain this difference. Qualitative differences in the design of the lifting work, with more lifts of long duration in awkward work postures, more carrying, more exertion of horizontal forces and a greater frequency of unexpected rapid lifts in the nursing aides may also contribute to the differences in risk of injury between these occupational groups. Considerable differences were found between the two hospital wards as regards lifting frequency, force, duration and the proportion of lifts with an even distribution of load on both feet during upward lift. The lifting work was approximately 50% less in the modern ward, which has easily maneuvered electric overhead hoists, spacious premises and better work organization than in the traditional ward, even though the patient handling needs were equivalent. In the warehouse department at which the wares were stored with easy access, the frequency of lifts with an even distribution of load on both feet was approximately 25%, compared to 7% in the other warehouse department. In conclusion, evaluations of occupational lifting work should include a quantitative assessment of the duration and load distribution of lifting, as well as of postures. Workstation design, technical aids and work organization have a profound effect on the lifting work load even in jobs with equal manual handling demands.