SummaryAn experiment designed to assess the components of variability of a number of measures of platelet aggregability showed that the ADP ED50 (the estimated dose of adenosine diphosphate at which primary aggregation occurs at half its maximum velocity) had the least method error of any of the parameters measured, but that none had a very high between-person component of variability. A simultaneous comparison of a syringe technique and a free-flowing technique for venepuncture revealed no differential effects on the aggregation parameters measured. An enforced increase in the stirring speed in the aggregometer led to an experiment which showed that such a change did not apparently affect the ED50s.