A double masked, placebo controlled clinical trial, of topical 2% ketoconazole cream with lid hygiene, for the treatment of seborrhoeic and mixed seborrhoeic/staphylococcal blepharitis was conducted. Forty patients with symptomatic blepharitis, 20 randomly allocated to ketoconazole, 20 to placebo were entered. Treatment efficacy was assessed by improvement of symptoms using visual analogue scales, appearance of the eyelids and reduction in numbers of pityrosporum yeasts on the eyelids. Both groups showed a similar reduction in symptoms, and signs of inflammation. The overall clinical impression in the ketoconazole group was better at week five than in the placebo group. Sixty nine per cent were either normal or markedly improved, as compared to 42% in the placebo group, although this was not statistically significant (p less than 0.1, one-sided Mann-Whitney). Pityrosporum numbers were reduced significantly in both groups during the treatment period. Ketoconazole was no better than placebo at improving the symptoms of blepharitis. More ketoconazole treated patients had normal or markedly improved lids after treatment than the placebo group. Pityrosporum yeasts may play a role in blepharitis, and treatment with an antifungal has some advantages over conventional therapy.