The proliferative status of thymocyte cell subsets in vivo was assessed by observing the effect of two antimitotic drugs, hydroxyurea (HU) and demecolcine. Both drugs had the greatest effect on the double-positive (DP) subset followed by the L3T4+ single-positive (SP) subset. However, the decrease in the latter type was delayed by several days, showing that their precursors rather than the cells themselves were killed by HU. Double-negative (DN) cells were less affected, indicating that they contain a resting subset and that they are renewed by emigration rather than by autonomous in situ proliferation. After drug treatment all cycling cells were eliminated from the thymus but, as shown by in vivo and in vitro bromodeoxyuridine incorporation, new cells rapidly reentered in cycle, starting from DN cells and Lyt-2+ SP cells and followed by DP cells. Lyt-2+ SP cycling cells represent an intermediary stage between DN and DP cells, and they are very transient. Injection of HU 24 h after in vivo BrdUrd labeling eliminated most labeled DN cells, but did not prevent the emergence of L3T4+ SP-labeled cells on day 3 as observed in control thymuses. These results suggest that these L3T4+ SP cells are generated from DP cells in the absence of proliferation. Cycling cells in the regenerating thymus were first located at the corticomedullary junction and then in the subcapsular region, suggesting a reverse migration process to that observed after cessation of proliferation. A model is proposed to summarize these sequential events.