Human lymphocytes were assessed as a cellular model for determining the effects of drugs on human mitochondria. Formation of total oxidized 14C-products was maximal with 1 mM [U-14C]palmitic acid, was linear for 90 min, linear with the number of lymphocytes, and decreased by 95% and 77% in the presence of 30 microM rotenone and 2 mM KCN. Seven drugs were tested which had previously been shown to inhibit beta-oxidation in animals; all decreased formation of total oxidized 14C-products by human lymphocytes, but with different IC50 values: 35 microM with amiodarone, 2.75 mM with tetracycline and amineptine, 3.75 mM with tianeptine, and more than 10 mM for valproic acid and the ibuprofen enantiomers. Formation of [14C]CO2 either increased or decreased, in relation to the various effects of these drugs on coupling, beta-oxidation, and the tricarboxylic acid cycle. There was a general trend for some relationship between inhibition of fatty acid oxidation and loss of cellular ATP. Those compounds, however, which uncoupled oxidative phosphorylation (2,4-dinitrophenol, amiodarone, ibuprofen) and/or inhibited the mitochondrial respiratory chain (amiodarone, rotenone, KCN) resulted in comparatively higher ATP depletion. Amiodarone, a drug which produces several effects (uncoupling, inhibition of beta-oxidation, of the tricarboxylic acid cycle and of the respiratory chain), caused a dramatic decrease in cellular ATP and cell viability at low concentrations (20-100 microM). Both these effects were prevented by the addition of 5 mM glucose, a substrate for anaerobic glycolysis. We conclude that human lymphocytes may be a useful model for assessing the effects of drugs on human mitochondrial function. IC50 values determined with this model may not necessarily apply, however, to other cells.