Pilot studies can be used to identify adequate test formulations for pivotal bioequivalence trials. The objective of this study was to evaluate the usefulness of pilot studies in predicting ratios and the intrasubject coefficient of variation (CV w ) for pivotal studies of high-variability drugs. Seven cross-over and replicate bioequivalence trials were selected. A hundred simulations of pilot studies were performed for different sample sizes and designs. The pharmacokinetic data of the selected formulations were analysed using WinNonLin based on an analysis of variance (ANOVA). The CV w was estimated using the formula recommended by the European Medicines Agency based on the mean square of the ANOVA. We calculated the predictivity index AE 10% and AE 20% of the real value. The predictivity index of AE 20% in the 2 9 2 design with 12 volunteers was 100% for AUC 0-t ratio, 87% for C max ratio, 50% for the CV w of AUC 0-t and 52% for the CV w of C max . The results of the 4 9 4 design with 8 volunteers were similar to those of the 2 9 2 design with 12 volunteers. These results were worse for the predictivity index of AE 10% in both designs. Pilot studies do not seem useful for predicting sample size. However, they were very good for predicting the AUC 0-t ratio and good for predicting the C max ratio. The most adequate design for pilot studies seems to be the 2 9 2 design with at least 12 volunteers.In the development of generic drug products, pilot trials are used to identify adequate test formulations for pivotal bioequivalence trials [1,2]. The Unites States Food and Drug Administration recommends that such pilot studies are carried out in a small number of volunteers before conducting the pivotal study of high-variability drugs [2]. Consistent with Arain et al., we can consider a pilot study to be a 'small study for helping to design a further confirmatory study ' [3,4].Pilot studies are mainly used to collect information on aspects such as the validation and determination of the analytical method, the assessment of the variability of the drug and the optimization of sampling times in the determination of pharmacokinetic parameters [2,3].The results of pilot studies are rarely published in scientific journals [3,5,6], and the literature provides no formal methodological guidance for conducting such studies [5]. However, it is true that authors such as Fuglsang [7], Mathew et al. have studied different methods and tools for the development of pilot studies. Standardization of these procedures could ensure appropriate design and development of pilot studies, which would in turn enable correct evaluation of the drug and successful design of the pivotal study.In pivotal studies, the test and reference formulations can be considered bioequivalent if the 90% confidence interval of the ratio of the means of the pharmacokinetic parameters (area under the plasma concentration curve from administration to the last observed concentration at time t [AUC 0-t ] and maximum plasma concentration [C max ]) is within ...