Several models for low density lipoprotein (LDL) apo B metabolism were applied to LDL turnover data from subjects with two distinct genetic forms of hyperlipidemia, familial hypercholesterolemia (FH), and familial combined hyperlipidemia (FCHL). Of the first two models tested, there was good agreement between the observed and predicted data for FH in one (model A), and for FCHL in the other (model B). The major difference between these two models is that LDL is kinetically homogeneous in model A and heterogeneous in model B, raising the possibility that LDL subspecies differences may occur between these two disorders. The findings are consistent with LDL homogeneity in FH and LDL heterogeneity in FCHL. Two other integrated models (models C and D) provided good agreement between observed and predicted data in both disorders. Although neither could be rejected outright on the basis of known physiology, parameter estimates were more variable with model D. Analysis of the data using model C was consistent with the known pathophysiologic defect in LDL catabolism in FH and suggests that individuals with FH as well as FCHL have more than one LDL subpopulation in plasma. The urine/plasma (U/P) ratio was shown to be constant from day 4 to day 14 of the study in FH, while in FCHL this value declined in all cases. Thus, determination of LDL fractional catabolic rates (FCR) by the U/P ratio method may be invalid in certain groups of patients. The other traditional method for calculating LDL FCR, the Matthews' analysis, overestimated FCR in some instances, and could lead to systematic errors when used to determine LDL FCR and production rates.(ABSTRACT TRUNCATED AT 250 WORDS)