It is now well known that the vasoconstrictor substance in platelets is 5-hydroxytryptamine (serotonin) (Rand & Reid, 1951). Humphrey & Jaques (1954) have estimated the amounts of serotonin present in platelets of different species of animals, and have found that this substance is present in larger amounts than histamine. It is not known why platelets should contain such large amounts of serotonin. In this paper it will be shown that platelets from dogs are capable of taking up serotonin from solution. This finding raises the possibility that the serotonin found in platelets may be accumulated rather than formed in situ. METHODS Dogs were anaesthetized with phenobarbitone. A short glass cannula, previously treated with silicone, was inserted into the femoral artery and joined to a short length of polythene tubing for collection of blood.Preparation of platelets. Blood was collected into silicone-coated vessels and mixed with 7&th vol. of 1 % 'Versene' (disodium ethylenediaminetetra-acetic acid) in 0 7 % NaCl solution. After centrifugation for i hr at 1600 g in a refrigerated centrifuge, the buffy layer was collected and transferred to narrow silicone-coated centrifuge tubes, in which it was again centrifuged for J hr at 1600-2000 g. The top portion of the buffy layer could now quite easily be collected by means of a pipette coated inside and out with paraffin wax, and consisted almost entirely of platelets. There were few or no red cells, and the proportion of white cells to platelets was about 1:500-1:1000. The platelets were suspended in 10 ml. of the fluid described below, and centrifuged at 40 C for J hr at 2000 g. The supernatant (which contained negligible amounts of serotonin or histamine) was discarded and the washing repeated. The platelets were then suspended in the suspending fluid, and their number counted, using a haemocytometer. When stored at 40 C they were quite stable, showing no tendency to agglutinate or to lose their contained serotonin into the surrounding fluid over periods of 5 hr.Preparation of red cells: Lys8i ofplatelets. Red cells, free from platelets, were washed thrice with suspending fluid by centrifugation at 40 C. It had been observed in other experiments (Humphrey & Jaques, 1954) that rabbit platelets which were alternately frozen and thawed released into the surrounding fluid any serotonin or histamine which they contained. The amounts of histamine released in this way were, within experimental error, the same as could be extracted by brief * Singapore, Colonial Development and Welfare Scholar. Present address: Physiology Department, University of Malaya, Singapore.