No dermatophytic hyphae was found. Culture of these specimens yielded no dermatophytic growth. A local mycology laboratory confirmed our findings. His complete blood picture, fasting glucose, and liver and renal function tests were normal. Human immunodeficiency virus antibodies and HBsAg were negative.Our provisional diagnosis was M. furfur infection affecting his trunk, palms and fingernails. We prescribed oral itraconazole 200 mg daily for 1 week. The patient attended follow-up 3 weeks after completion of such therapy. Little remission of his cutaneous and nail lesions and no improvement in his pruritus were noted.We prescribed oral itraconazole 200 mg daily for one more week. Almost total ablation of pruritus, truncal and palmar lesions was noted another 3 weeks later. Fingernail dystrophies became more distal, with fresh nail plates noted from the matrix end.Our case and other reports 1-3 substantiate that M. furfur can cause onychomycosis. Whether M. furfur is the pathogenic culprit or merely an innocent bystanding coloniser remains debatable, as keratolytic properties of this pathogen were not well established. However, distal subungual dermatophytic onychomycosis involves the nail bed only in the initial phase. 4 The absence of keratolytic properties therefore does not preclude M. furfur from causing onychomycosis.For non-dermatophytic filamentous fungi, a 'simple association' evaluation criterion -Wood's light fluorescence, positive microscopic results for hyphae and spores and negative dermatophytic culture -incurs high specificity in determining their pathogenic significance. 5 More stringent criteria also exists for onychomycosis caused by non-dermatophytic moulds. 6,7 The applicability of these criteria to Malassezia spp. onychomycosis in our case is unknown.From a study on 500 patients with PV reported by one of us (VZ), 8 many patients had onychomycosis demonstrating goldenyellow fluorescence under Wood's light. Their nail clippings yielded no dermatophytic growth. Cutaneous and nail lesions responded to systemic ketoconazole (itraconazole was not available then). Two patients had palmar Wood's light fluorescence. Cutaneous PV and palmar fluorescence disappeared after systemic treatments. 8 Persisting nail fungal reservoirs account for treatment failures for dermatophytic cutaneous infections. Longer courses of systemic terbinafine or itraconazole pulses are effective for treating these patients. Treatment failures for PV frequently occur despite systemic itraconazole. 8 We thus postulate that severe pruritus in widespread M. furfur infection may cause frequent scratching, leading to M. furfur onychomycosis. M. furfur nail reservoirs might then cause treatment failures. We therefore advocate open case-control studies, and subsequent double-blind randomised trials, to investigate whether longer treatment courses or pulse therapy can cure patients with PV and M. furfur onychomycosis. Onychomycosis: a critical study of techniques and criteria for confirming the etiologic significance of nondermatop...