We have confirmed our earlier finding that most red wines are able to bring about 5-hydroxytryptamine (5-HT, serotonin) release from platelets in vitro. Platelets from individual subjects manifested varying degrees of releasing ability but responded to different wines with a similar rank ordering. There was a high correlation (r = 0.87) between the effect of red wine and that of reserpine in different individuals. Some types of red wine caused a consistently higher release of 5-HT than others in all subjects; one red wine in particular resulted in negligible release. When several brands of this 'low-releasing' red wine were further examined, they all showed a lower activity than all the brands of a 'high-releasing' red wine type. This variation in releasing power was not related to intensity of red colour. Partial purification of red wine was achieved by column chromatography and showed releasing activity to be associated with a low molecular weight orange fraction. Preliminary studies, using solid phase extraction methods, showed that the active components lie mainly in a subgroup of the flavonoid fraction. If any of the adverse effects of red wine, such as headache induction, derive from this 5-HT releasing ability, then it may be possible to prepare red wines free from the chemical substances responsible.