Our specific aim was to assess within-family clustering of high-density lipoprotein cholesterol (HDLC) levels in kindreds identified through probands with primary hypoalphalipoproteinemia, and to determine whether, and to what degree, familial aggregation of HDLC less than or equal to the tenth percentile represents a heritable trait, familial hypoalphalipoproteinemia. Our probands were selected arbitrarily by virtue of HDLC less than or equal to the age-sex-race-specific tenth percentile as the sole dyslipoproteinemia, with an additional requirement that they be normotriglyceridemic (triglyceride levels less than the 90th percentile). The probands were also required to have primary hypoalphalipoproteinemia, not secondary to diseases and/or drugs. Fifteen of the 16 probands were men; 12 were referred because of premature myocardial infarction, angina, or stroke, 2 because of family history of premature myocardial infarction or stroke, and 2 because of low HDLC observed on routine health examinations. Two of the 16 kindreds exhibited three-generation vertical transmission of bottom decile HDLC. In three kindreds, there was also three-generation vertical transmission of bottom decile HDLC, but top decile triglycerides accompanied bottom decile HDLC in one or more generations. Eight kindreds displayed two-generation vertical transmission of bottom decile HDLC. After excluding probands, there were 11 critical matings (bottom decile HDLC by normal), with 30 living offspring, all of whom were sampled. Of these 30 offspring, 13 had bottom decile HDLC, 17 had HDLC greater than tenth percentile. The ratio of offspring with bottom decile HDLC to those of HDLC greater than tenth percentile was 13:17 (0.76/1), not significantly different from the ratio of 1/1, the ratio predictive of a dominant trait, X2(1) = 0.53, P greater than 0.4. The nearly 1:1 segregation ratio for the group of offspring was not due to the aggregation of sibships with, in general, most of the sibs, or none of the sibs affected; within-family expression of low HDLC was also not sex-linked. The 13 hypoalphalipoproteinemic offspring of 11 critical matings included only two subjects whose bottom decile HDLC was accompanied by top decile triglyceride. Our data suggests that not only (by selection) was low HDLC in the probands the sole dyslipoproteinemia, but that the segregation of low HDLC in offspring of critical matings was primarily accounted for by isolated low HDLC, not by hypoalphalipoproteinemia secondary to hypertriglyceridemia. Familial hypoalphalipoproteinemia is a heritable disorder with a pattern of transmission not significantly different from that expected by a hypothesis of mendel