Purpose: Results from in vitro chemosensitivity testing recommend treosulfan/gemcitabine chemotherapy for the treatment of stage IV uveal melanoma. Methods: Twenty patients received treosulfan 3,500 mg/ m2 followed by gemcitabine 1,000 mg/m2 on day 1 and day 8 repeated on day 29. In cases of prior chemotherapy only 75% of these dosages were used. Results: Without any patient achieving an objective response, 25% of patients (95% confidence interval, 8.6-49.1%) had stabilisation of disease. This stabilisation was associated with a prolonged median overall survival of 17 months compared with 7 months for the patients with progressive disease. First-line treatment was not associated with better response or survival although prognostic parameters did not significantly differ from that of other patients. Conclusions: The results are disappointing and question the value of individualized chemotherapy based on in vitro assays.