The extent to which bone marrow obtained by conventional aspiration is contaminated by peripheral blood has been confirmed and quantitated. In marrow aspirates from normal subjects the median percentage of nucleated cells that had originated from the peripheral blood was 32% (range 2.5%-64%), in patients with acute leukemia 23% (range 0.5%-96.5%), in patients with chronic leukemia 59% (range 17%-76%), and in patients with lymphoma 31% (range 0.5%-74%). Flow cytometric (FCM) DNA analysis of conventional marrow aspirates from a range of subjects significantly underestimated the proportions of S-phase cells present, when compared with results from trephines obtained at the same time. Having shown, using 51Cr-labeled red cells in mice, that circulating red cells do not reenter the marrow parenchyma, a mathematical correction for contaminating blood similar to that described by Holdrinet et al. was devised. This correction improved the S-phase cell estimate from aspirated marrows, and the corrected values were not significantly different from values from paired trephine samples. A previously described technique for collecting fragments by filtration of aspirated marrow has been adapted for FCM analysis as a more direct way of overcoming problems due to blood contamination. This method was shown to yield estimates of S-phase cells not significantly different from those in paired marrow trephines and offers an alternative to routine trephine biopsies for FCM analysis of marrow cell kinetics.