The epidemiological characteristics of platelet aggregability were established in 958 participants in the Northwick Park Heart Study. The main analyses were based on the dose of adenosine diphosphate at which primary aggregation occurred at half its maximum velocity. Aggregability increased with age in both sexes, was greater in whites than blacks (particularly among men), and tended to decrease with the level of habitual alcohol consumption. Aggregability was, however, greater in women than men and in nonsmokers than smokers. There was no relation between aggregability on the one hand and obesity, current or past oral contraceptive use, menopausal state, or blood cholesterol and triglyceride concentrations on the other. Aggregability was somewhat, though not significantly, higher in men with a history of ischaemic heart disease and in those with electrocardiographic evidence of ischaemia than in those without. There was a strong association between the plasma fibrinogen concentration and aggregability. The widely held concept of platelet aggregability and its implications is probably an oversimplification. In the prevention of thrombosis it may be as useful to consider modifying external influences on platelet behaviour, such as plasma fibrinogen concentration or thrombin production, as it is to rely solely on platelet active agents.