Background: Onychomycosis is a difficult condition to treat and cure rates are disappointing. Moreover fungicidal action of antifungal agents in NCCLS assays and their rapid accumulation in nails in vivo are not compatible with the duration of treatment. Aims: This study aimed to find the effectiveness of 4 different antifungal agents in an in vitro model with some similarities to in vivo conditions. Materials and Methods: Strains of Trichophyton rubrum I-III, Trichophyton mentagrophytes (usual form), Trichophyton mentagrophytes 73, Epidermophyton Flucosom, Microsporum Canis, and Trichophyton Schoenleini which were isolated from the nails of patients, were hired. Inocula suspensions were prepared from 7 to 14 day-old cultures of dermatophytes. Antifungal agents including fluconazole, ketoconazole, terbinafine, and griseofulvin were obtained as standard powders. For each antifungal agent, initial MIC was calculated by registering the optical density for 10 two-fold serially diluted forms which was incubated with diluted fungal suspensions with RPMI 1640. Human nail powder inoculated with different strains and incubated in RPMI 1640 and different concentrations of antifungal drugs for 4 weeks. Final MIC at different steps of 1, 2, 3 and 4 weeks were investigated. Results: The final MIC that resulted from the incubation of dermatophytes with nail powder was much more than the initial which was concluded from conventional MIC assay. Terbinafine had the lowest rate of initial and final MICs. Conclusion: The model described here may present more similar conditions to clinical fungal infections; therefore the results such as MIC may be more helpful for hiring the most effective antifungal agent.