OBJECTIVE: The aim of this study was to test the bioequivalence of two alendronate tablets (CAS 121268-17-5; Marvil 10 and Marvil 70 as test formulations, in short "test"; reference formulation, in short "reference") in vitro and in vivo in healthy adult male subjects and to describe a mode for researching the bisphosphonate oral formulation pharmaceutical quality. METHODS: Two dissolution tests with 10-mg and 70-mg alendronate tablets, a preliminary clinical test with 10-mg tablets (n = 10) and a bioequivalence study with 70-mg tablets (n = 23) were performed. Clinical studies were single-dose, open, cross-over, randomized, including a four-week wash-out period. Alendronate was assessed by HPLC in urine after 6 (UE6) and 24 (UE24) h post-intake. In all the experiments the reference was the one that had proved efficacy and safety in international regulatory clinical trials. RESULTS: The dissolution test showed a comparable release profile between reference and test, of both, the 10-mg and 70-mg tablet, the difference (f1) and similarity (f2) factors being within the acceptance values. The clinical trials showed great variability in urinary recovery, from one third the average figure up to 2-3 fold. The amount recovered with the 70-mg tablet was 11-15 fold higher than with the 10-mg tablets, suggesting higher (test/reference) was found to be 72-122% for UE24, and when analyzed in individuals with apparent steady bone metabolism during the wash-out period (n = 19), it was 86-137%. Both margins are considered acceptable in view of the particular kinetic and dynamic features of bisphosphonates, their very high inter- and intra-individual variability, extremely low absorption, time-changeable bone compartment, high margin of safety and long-term achievable therapeutic benefits. CONCLUSION: Test is bioequivalent to reference.