In Reply I appreciate the opportunity to respond to a letter from Kivelä and Finger regarding our recent article.1 This article confirmed prior reports showing that the prognostic accuracy of gene expression profiling (GEP) is superior to that of anatomic features used by the TNM staging system for uveal melanoma,2,3 and it showed that class 2 uveal melanomas were less likely to metastasize when their largest basal tumor diameter was less than 12 mm. The main point of the article was that tumor diameter should be taken into account when planning metastatic surveillance and clinical trial enrollment of patients with small class 2 uveal melanomas.