Objectives-To determine the value of using routinely collected sickness absence data as part of a health needs assessment of healthcare workers. Method-Sickness absence records of almost 12900 NHS staV for one calendar year were analysed. Three measures of absence, the absence rate, the absence frequency rate, and the mean duration of absence, were assessed for the population and comparisons made between men and women, full and part time and diVerent occupational groups of staV. Also, the main causes of sickness absence were found. Results-Almost 60% of the study population had no spells of sickness absence in the year of study and almost 20% had only one spell of sickness absence. Female staV were more likely to have experienced sickness absence than male staV. Although absence due to conditions related to pregnancy were included in the analysis, the incidence of these was not suYcient to account for the higher rates of absence among female staV. In general, full time staV had greater rates of sickness absence than part time staV. 71% of all absences were of <1 week duration. The main known causes of sickness absence were respiratory disorders, digestive disorders, and musculoskeletal disorders. Conclusions-The transition from units managed directly from the health board to trusts with individual responsibility for personnel issues at the time of data collection resulted in variations in the quality of data available for analysis. This together with the use of "dump" codes has influenced the quality of the analysis. However, such data should be available for analysis to tailor occupational health care to the needs of the population. (Occup Environ Med 1999;56:702-708)