Of 414 recorded hospitalizations among British troops during Operation Joint Endeavor (Bosnia) in 1996, 2% were attributable to battle injuries, 46% to routine injuries, and 52% to disease; most injuries were to the lower limbs, and most diseases were of the skin and musculoskeletal system. The median length of inpatient stay was 4 days for injury (range, 1-146 days) and 3 days for disease (range, 1-60 days). Correcting for uncaptured data, the number of hospitalizations attributable to both injury and disease was significantly lower than the number predicted from the NATO planning figures (p < 0.0001). The NATO planning estimates for expected hospitalizations need to be revised. The electronic recording at source of all patient information should be introduced during military deployments to optimize data capture, facilitate the audit of clinical activity, and inform future medical planning. All hospitalizations occurring during military missions should be recorded, and surgical interventions during deployments should be coded at source. Ineffective paper-based morbidity surveillance procedures such as Jefferson 97 (or J97) must be discontinued and replaced by an electronic, fully integrated, NATO-wide clinical information system. This clinical information system should encompass primary care, secondary care, medical training, and medical supply. It should be built on a Microsoft Office platform and should be capable of interfacing electronically with existing civilian databases. Important clinical outcomes should be structured hierarchically within the data set. The database should be configured so that it can be accessed locally by clinicians and remotely by epidemiologists and planners.