The efficiency and efficacy of low-density lipoprotein (LDL) apheresis performed with a dextran sulphate cellulose (DSC) regenerating unit were tested in five familial hypercholesterolaemic patients. LDL apheresis was repeated four times at both bi-weekly and weekly intervals, processing one plasma volume each time. The efficiency of the procedure (i.e., the extent of lipoprotein removal) was nearly identical with both schedules. Efficacy parameters, i.e., decreases of plasma total and LDL cholesterol (TC and LDL-C) and apo B, were highly correlated (r greater than 0.96) with preapheresis levels, allowing an accurate prediction of the absolute lipid removal in the single individual. Plasma triglycerides, high-density lipoprotein cholesterol, apo A-I and apo A-II recovered rather rapidly, reaching 91-96% of the pre-apheresis values in 48 hours; the recovery of TC, LDL-C and apo B was much slower, with a relatively rapid early phase (80% recovery after about 7 days) followed by a successive slower rise. This pattern was highly reproducible in the single patient, allowing the definition of a simple mathematical model for an accurate (error less than 20%) prediction of the individual process. Based on this model one can design the treatment schedule necessary to maintain lipid levels within the desired range in the single individual. The hypolipidaemic efficacy of DSC apheresis appears, otherwise, not to be dependent upon the procedure per se, but on other individual factors, e.g. the amount of removable lipoproteins and the rate of lipid recovery; both can be predicted with sufficient accuracy.