Background: Data about relapses of onychomycosis after treatment with the new systemic antifungals vary among the different studies, with figures ranging from 3 to 20% for terbinafine and from 21 to 27% for itraconazole, depending on the follow-up duration. Objective: To determine the prevalence of relapses of onychomycosis cured by terbinafine compared with that of onychomycosis cured by itraconazole. Methods: We followed up 47 patients whose toenail onychomycosis had been mycologically cured in an open randomized study comparing intermittent itraconazole treatment with continuous terbinafine treatment and intermittent terbinafine therapy. Patients were examined every 3 months for up to 3 years after the end of therapy. At each visit clinical and mycologic (direct microscopy and cultures) evaluations were performed. Results: Eight of the 36 patients (22.2%) who completed the study had a relapse of onychomycosis during the follow-up period, including 2 patients of the terbinafine 250 mg group, 2 patients of the terbinafine 500 mg group and 4 patients in the itraconazole 400 mg group. As the original infection, the relapse was caused in all cases by Trichophyton rubrum. Conclusions: This study shows that 22.2% of patients with onychomycosis successfully treated with systemic antifungals experienced a relapse. The relapse rate increased from 8.3% at month 12 to 19.4% at month 24 and to 22.2% at month 36. Relapses were more common in patients treated with pulse itraconazole (4/11) than in patients treated with continuous (2/12) or intermittent (2/13) terbinafine. Statistical analysis did not reveal any significant difference between relapse rates in the three groups.