The familial aggregation of lipids [total cholesterol (CH) and triglyceride (TG)] and lipoproteins [high-density lipoprotein cholesterol (HDL) and low-density lipoprotein cholesterol (LDL)] was investigated in families ascertained through both random and nonrandom probands in the Minnesota Lipid Research Clinic Family Study. Nonrandom proband ascertainment was based on single selection through truncation for hyperlipidemia at an earlier screening. A path model was used to investigate the nature of familial resemblance using appropriate adjustments for ascertainment and to determine whether random and hyperlipidemic samples are heterogeneous with regard to the multifactorial model. The results suggest that parameter estimates are consistent with those from previous studies in which only random families were used and that random and nonrandom samples are homogeneous with regard to the path model for CH and LDL. However, for TG and HDL the random and hyperlipidemic samples are significantly heterogeneous. This heterogeneity would be observed if familial hypertriglyceridemia and/or familial hypoalphalipoproteinemia segregates predominantly in the hyperlipidemic rather than in the random sample, as on might expect.