The three classical pharmacokinetic parameters used to assess bioequivalence, AUC (total area from zero to infinity), Cmax (peak plasma concentration), and tmax (time to reach Cmax), are suitable to determine the extent and rate of absorption of immediate-release drug products. However, they may fail to evaluate the pharmacokinetic performance, particularly the rate of absorption of sustained-release (SR) formulations, which yield flat plasma curves with multiple peaks. This paper evaluates the inclusion of the following criteria for bioequivalence assessment of diltiazem SR formulations: MRT (mean arithmetic time), Cmax/AUC, peak occupancy time (POT), t(apical) (the arithmetic mean of the times associated with the concentrations within 25% of Cmax), C(apical) (the arithmetic mean of the concentration within 25% of Cmax), percent fluctuation and flatness of the curve as assessed by the coefficient of variation of the Css (steady state concentration) values obtained during a dosing interval at steady state. The above proposed criteria, as well as the classical parameters AUC, Cmax, and tmax were utilized in a recent pharmacokinetic study of a new SR product of diltiazem, Dilapress 240 (formulation A). Formulation A was analyzed following single (240 mg) and multiple (240 mg qd for 6 days) dosing at steady state (day 6) in comparison to Cardizem CD (formulation B). The bioavailability of formulation A relative to that of formulation B following single and multiple dosing was 92 +/- 28% and 90 +/- 24%, respectively. The 90% confidence intervals (Cl) over a mean AUC ratio of 89% were 78-101% (single dose, SD) and 77-101% (multiple dose, MD). following the administration of formulations A and B, identical mean values of the peak plasma concentration were obtained: 84 ng/mL (SD) and 132 ng/mL (MD). The 90% Cl over a mean Cmax ratio of 100% were 83-115% (SD) and 86-115% (MD). In the SD study, subject 8 had a relative bioavailability value of 24%, which deviated by 7.5 standard errors (SE) from the mean AUC ratio. Consequently, we repeated the single dose analysis without subject 8. The mean bioavailability data was 97 +/- 37% with a 90% Cl of 80-114% over a mean AUC ratio of 92%. ANOVA analysis did not show any formulation or period effect in all tested pharmacokinetic parameter s. On the basis of these results, these two formulations were judged to be bioequivalent. In contrast to the AUC and Cmax ratio, the 90% Cls associated with the ratio of the proposed criteria, with the exception of C(apical), did not fall within the acceptable limits. In the current study, a discrepancy was found between the above pharmacokinetic parameters, which were examined concerning their ability to detect differences in bioequivalence between SR products and the classical parameters regularly used for bioequivalence assessment. Although the parameters examined are theoretically more attractive than the single point parameters Cmax and tmax for rate of absorption assessment, their utility in bioequivalence would require further examination.