Diagnostic accuracy for melanoma was determined in a dedicated pigmented lesion clinic. We assessed the impact of duration of experience in dermatology and also the relationship between tumour thickness and accuracy of clinical diagnosis. We reviewed the histopathology request forms and reports for all biopsies generated by the Pigmented Lesion Clinic, Western Infirmary, Glasgow during 1992-94 inclusive. The clinic is staffed by two consultants, one senior registrar and one registrar. Diagnostic accuracy, index of suspicion, sensitivity, specificity and positive predictive value were calculated for the clinic overall, and for each grade of staff. One hundred and sixty-three lesions were diagnosed clinically as melanoma. A histopathological diagnosis of melanoma was made for 128 lesions during this period, 113 of which had been correctly diagnosed before surgery. The diagnostic accuracy for two dermatologists each with > 10 years experience in dermatology was 80%, with sensitivity of 91% and positive predictive value of 86%. Diagnostic accuracy rates for two senior registrars (each with 3-5 years experience) and six registrars (each with 1-2 years experience) were 62% and 56%, respectively. Thin and intermediate thickness melanomas generated the greatest inaccuracy irrespective of clinical experience, although registrars failed to recognize melanoma three times more often than the other groups. We report the diagnostic accuracy for melanoma by trained dermatologists to be higher than previously reported. In comparison with trainees, > 10 years experience in dermatology and exposure to more than 10 melanomas per year appears to be associated with greater diagnostic accuracy. Knowledge of the current clinical diagnostic accuracy at varying levels of experience is essential if the impact of training is to be evaluated. As pigmented lesions of virtually all types can be treated within dermatology departments, dermatologists are the appropriate first point of referral for suspected early melanoma.