Abstract
Treatment of uveal (intraocular) melanoma is aimed at prolonging life, if possible conserving the eye and useful vision. About 50% of patients develop fatal metastatic disease despite successful eradication of the primary intraocular tumour. The effect of ocular treatment on survival is unknown, because the same survival data from case series can be interpreted in different ways. Treatment is therefore based on intuition and varies greatly between centres. Randomised trials of treatment vs non-treatment of asymptomatic tumours are desirable but would be controversial, difficult, expensive and possibly inconclusive. Strategies for coping with uncertainty are needed to avoid unethical care.