A $3{ imes}3$ Latin square crossover study for the bioequivalence of three ondansetron formulations was conducted. Test products were $Vominon^{circledR}$ 8 mg and $Vominon^{circledR}$ 4 mg tablets and reference product was $Zofran^{circledR}$ tablet. Twenty one healthy Korean male subjects received each formulation at the ondansetron dose of 8 mg and plasma concentrations of ondansetron were monitored by HPLC for over a period of 12 hr after the oral administration. Statistical procedure for bioequivalence evaluation of AUC {e.g., analysis of variance (ANOVA), multiple comparison and confidence intervals} was carried out. There were no significant differences in AUC among the formulations. The confidence intervals for the AUC of $Vominon^{circledR}$ 8 mg and $Vominon^{circledR}$ 4 mg were between -0.24 and 15.54% and between -2.41 and 13.36% respectively, within a range that proposed by the Korea Food and Drug Administration Guidelines for Bioequivalence. These statistical procedure could be standardized and generally applicable for the assessment of bioequivalence for multiple (more than two) formulations.