Fifteen bone marrow (BM) and venous blood circadian profiles were obtained from 13 diurnallyactive, healthy men sampled every 4-5 h for 24 h. Peripheral blood (PB) was also sampled in subsets of 5 men either for 24 h immediately preceding the BM procedure or 5-6 months afterwards. Cortisol and white and red cell parameters were determined in PB. BM cell cycle distribution was investigated in parallel by flow cytometry for S-phase DNA of total mononuclear cells and subpopulations of erythroid and myeloid precursor cells. On a group basis, significant circadian rhythms were found in PB variables commonly referred to as "marker" rhythms (cortisol, total white cells [WBC], neutrophils [N], lymphocytes [L]), with acrophases less than 2 h apart between the control day prior to and during BM sampling. Thus, major, but relatively short-lasting physiological stress, like BM aspirations or blood sampling itself, although repeated several times over 24 h, seemed to have minor influence on these rhythms on days of the BM procedure. When comparing the times of highest or lowest values in PB with times of highest or lowest values in BM, several temporal relationships were found. Among other associations, timing of lowest values in WBC, N and L or highest values in cortisol was significantly predictive of highest values in myeloid cells occurring in the following 12 h, whereas highest values in erythroid cells occurred significantly more often in the 12 h interval beginning 4 h after the time of lowest values in WBC, L and N. The stability in the circadian rhythms of the PB variables suggests that information obtained on one day can be used to guide procedures on the next, such as BM myelotoxic chronotherapy or BM harvesting.