We have examined megakaryocyte process fragments that migrate out of bone marrow explants after a short period of incubation and assume a beaded form, consisting of 2 or more putative platelets. The fragmentation appears to occur in vivo and supports the proposal that platelet liberation does not always occur in a sequential manner from the distal ends of megakaryocyte processes. Transmission electron microscopy revealed that microtubules were generally oriented longitudinally in the process fragments. Rarely, a microtubule coil was found in a terminally located putative platelet. The observations favour the view that the marginal coil of microtubules, which is a characteristic of circulating platelets, does not usually form until after platelets have been liberated.