The kinetics of low density lipoprotein apolipoprotein B (LDL apo B) metabolism are usually determined using turnover techniques in which radioiodinated LDL apo B is injected as a bolus into plasma, and serial plasma and urinary radioactivity samples are taken. The metabolic parameter of interest usually estimated from such data is the fractional catabolic rate (FCR). Two methods are normally employed to obtain an estimate of the FCR. One, the so-called Matthews' analysis, assumes plasma LDL apo B metabolism can be described by a single plasma pool while the other is determined by calculating the ratio of urinary radioactivity excreted to mean plasma radioactivity per day. Both of these methods assume LDL apo B is kinetically homogeneous, thus ignoring the evidence that LDL is biochemically heterogeneous in some individuals. If this biochemical heterogeneity manifests itself as kinetic heterogeneity, then the use of these data to estimate the FCR will not permit the resolution of the finer details of potential metabolic defects. This paper addresses the question of kinetic homogeneity and heterogeneity of LDL apo B within the context of several integrated kinetic models of increasing complexity. Each model fits reasonably the turnover data and hence cannot be rejected on the basis of failure to be compatible with the data. However, the models have strikingly different physiologic interpretations while providing essentially the same estimate for the FCR. Thus LDL apo B metabolism appears to be more complex than originally believed, and the models provide a framework within which to design new experiments to distinguish among them.