Since the detection that platelets originate from megakaryocytes (MK), the site of megakaryocyte fragmentation has been disputed. Some authors have even postulated that platelets are solely produced in the lungs. Thus, we have directly measured platelet generation in the bone marrow (BM) by comparing the relative number of young RNA-containing, so-called reticulated platelets (%RP) in the BM and in the peripheral blood (PB). Two separate prospective, cross sectional trials have been conducted in patients routinely undergoing BM biopsies for diagnostic purposes. In the first part of the study 30 patients with stem cell or bone marrow transplantation were examined. The second part of the study was performed in 62 haematological patients visiting the outpatient's clinic. Median %RP were higher in BM than in PB (p <0.001). In the second part of the study the difference averaged 133% (interquartile range: 30-383%). There was a moderate correlation between %RP in BM and in PB (r = 0.67; p <0.001). The absolute number of RP in PB correlated weakly with the number of megakaryocytes (0.42; p = 0.001), which was due to a correlation between the platelet counts and the megakaryocyte counts (r = 0.55; p <0.001 in biopsies). Two patients with autoimmune antibodies against GPIIb/IIIa exhibited 10% and 16% RP in PB, and had 29% and 59% RP in BM, respectively. It is concluded that the relative number of RP is significantly higher in BM than in blood. This supports the notion that platelets are at least in part released from MK in the bone marrow, particularly in patients suffering from immune thrombocytopenia.