Retirement may have a negative effect on a person's health, and may even accelerate death. To determine the validity of this effect, time intervals between reported retirement date and death date were tabulated for retirees from a large petroleum corporation. A regression analyses was performed to control for physiological factors, job status, and type of work differences that might influence longevity. Examination of the residuals showed a bimodal distribution: one group lived an average of 6.5 years beyond retirement date, while a second group lived an average of 19 years after retirement. The data suggest that short-term survivors after retirement may undergo a different adjustment process from long-term survivors. These results, combined with findings of other research, imply a need for further examination in the field of health care management. (Author/NRB)