BACKGROUND: The relative effectiveness of plaque radiotherapy and enucleation in patients with choroidal or ciliochoroidal melanoma remains controversial. Most published comparative survival studies have evaluated clearly biased groups, short-term follow-up, or both. METHODS: The authors performed a retrospective, non-randomized, comparative survival study on 734 patients with a choroidal or ciliochoroidal melanoma treated by enucleation (n=237) or plaque radiotherapy (n=497) between July 1980 and June 1985, and on a residual group of 448 patients, 103 treated by enucleation and 345 treated by plaque radiotherapy, identified by variable-by-variable range matching. Melanoma-specific mortality and all-cause mortality were evaluated by the life-table method. RESULTS: The melanoma-specific mortality rate was substantially worse in the original enucleation subgroup throughout available post-treatment follow-up; however, differences in baseline prognostic factors between the subgroups are likely to explain much of the survival differential. After eliminating patients with nonoverlapping values of individual clinical variables to adjust for recognized intergroup differences at baseline, there was no significant difference in the mortality curves of the residual subgroups. The cumulative 15-year survival based on deaths from metastatic melanoma in the enucleation subgroup was 57.1% (standard error=6.4%), while that in the plaque radiotherapy subgroup was 61.8% (standard error=3.3%). CONCLUSION: On the basis of these results, we conclude that a large difference in survival between equivalent groups of patients with primary choroidal or ciliary body melanoma treated by enucleation or by plaque radiotherapy appears unlikely.