Summary As part of a national campaign to combat the rising incidence of and mortality from cutaneous malignant melanoma, a programme of improved clinical services and professional and public education was set up in Nottingham in January to July 1987. The public education campaign in July led to an immediate increase in the weekly number of referrals to the pigmented lesion clinic from 10 to 54. The effect on general practitioner workload was less dramatic, the weekly number of consultations for discrete pigmented lesions rising from 0.5 to 3. In materials sent to GPs, we recommended that patients with three or more of seven specified signs should be referred for specialist opinion. Only 40% of the patients referred to the pigmented lesion clinic fulfilled this criterion, but 6% of these patients had a melanoma, compared to only 0.4% of those who did not meet the criterion. In the 6 months following the campaign, 64% of melanomas diagnosed in Nottingham residents had a Breslow thickness of <1.5mm whereas only four (16%) were >3.5mm. However, this distribution was not significantly different from that seen in the three and a half years before the campaign. These results suggest that attempts to improve early diagnosis of the disease by health education are justified, but, in view of the service implications, full evaluation of such campaigns by large scale and long-term studies is essential. Future campaigns should give greater stress to referral criteria.The incidence and mortality of cutaneous malignant melanoma in white populations is rapidly rising; mortality from the disease in England and Wales has more than doubled since 1950 (Osmond et al., 1983). Several large epidemiological studies show that melanoma is associated with exposure to ultraviolet light (Gallagher et al., 1986;Holman et al., 1986). However, the relationship is complicated, so as yet simple and effective primary preventive measures have not been developed.The best prognostic indicator for malignant melanoma is the Breslow thickness of the tumour, tumours with a depth of less than 0.76mm having a virtual 100% 5-year survival rate whereas tumours with a depth greater than 3.5mm have a 5-year survival rate of less than 40% (Breslow, 1970). The