[en] A major problem in occupational cohort studies is how to treat study subjects who are lost to follow-up (LTF). The assumptions made concerning their vital status may affect the results of comparative mortality analyses. The problem was considered within the context of an occupational follow-up study of white male employees at a nuclear facility in Colorado. In this analysis, 568 or 8% of cohort members were LTF. Comparative mortality for the entire cohort was estimated by treating LTF workers as lost at employment termination date, as living at the end-of-study date, and with cumulative mortality simulated between 0% and 100%. Results indicate that simulations of cumulative mortality among employees LTF can be useful in assessing the potential bias caused by LTF mortality assumptions. Further, a general method for assessing LTF bias in occupational analyses is proposed