The predictive value of melanoma diagnosis made by visual examination during pigmented lesion screening is low. This creates the problem of false-positive diagnoses, which lead to unnecessary treatment and scarring. The purpose of this study was to evaluate the impact of dermoscopy (epiluminescence microscopy, dermatoscopy) on the false-positive rate in the routine melanoma screening activity of a pigmented lesion clinic (PLC). In a series of 133 subjects consecutively referred to the PLC, lesions defined as suspicious or equivocal on visual examination were examined by dermoscopy. Only lesions also defined as suspicious on dermoscopy were excised; other lesions were observed at follow-up examinations. Among the 2542 pigmented lesions observed, clinical examination led to identification of 43 suspicious lesions, 13 of which were also suspicious on dermoscopy and were subsequently excised. Histopathological examination revealed three malignant melanomas. Compared with visual examination alone, the addition of dermoscopy to the subgroup of clinically equivocal lesions resulted in an increase in specificity from 98.4% to 99.6% and in positive predictive value from 6.9% to 23%. The specificity of the visit outcome 'subject to be referred for surgical excision' increased from 69.2% to 92.3%. No false-negative melanoma diagnoses on dermoscopy were observed during a follow-up period of 4 years. The addition of dermoscopy to routine PLC activity as a second-level examination led to a reduction in the number of false-positive diagnoses, thus producing an overall increase in the specificity and positive predictive value of melanoma diagnosis.