This study was designed to explore the incidence of lupus anticoagulant (LA) and anticardiolipin antibodies (ACA) and their relationship to each other in a healthy population of 499 blood donors. Plasma samples were tested for LA activity and IgG, IgM and polyvalent ACA. Prolongation of the kaolin clotting time of a mixture of 80% normal plasma and 20% test plasma compared to the normal (dKCT) was used to detect LA activity. A normal distribution of dKCT was found with the mean 3.5 seconds +/- SD 10.6 seconds. Forty subjects (8%) were greater than 10% of the normal control; among these, 18 (3.6%) were outside the 95% confidence limits. The median age (29.3) and sex (M = 12, F = 28) of the 40 subjects with prolonged KCT were significantly different (p less than 0.001) from the group as a whole, younger females predominating. The frequency distribution of IgG, IgM and polyvalent ACA was skewed and the majority did not have detectable levels. ACA concentration falling within 95% of the population group were regarded as normal. Applying this definition, abnormal IgG ACA was greater than 4.33 U/ml, IgM ACA greater than 3.55 U/ml and polyvalent ACA greater than 4.55 U/ml with a prevalence of 4.6%, 4.6% and 5.6% respectively. Of the subjects with positive ACA of any class there was no significant association with either age or sex or the presence of LA. Only three plasma samples had both activities. Neither ACA nor LA were associated with antinuclear antibodies (ANA) or rheumatoid factor (Rh factor). Thus, in a healthy population LA is found predominantly in younger females and neither LA or ACA appear to identify subjects with other autoimmune parameters such as ANA or Rh factor or, for that matter, each other.