Seventy-one patients receiving a minimum of two courses of high-dose intracavitary cisplatin (100-200 mg/m2/course) with i.v. thiosulfate administered to protect against cisplatin-induced renal insufficiency were retrospectively evaluated to examine the influence of thiosulfate and large cumulative doses of cisplatin on the incidence of hypomagnesemia. Only 8% of 50 patients who had normal serum magnesium levels prior to the initiation of the experimental program became hypomagnesemic during the therapeutic trial. Similarly, while 67% of the 21 patients with low initial serum magnesium levels remained hypomagnesemic, 33% had normal serum levels at the completion of therapy. It is suggested that the intravenous administration of thiosulfate might have been responsible for the strikingly low incidence of hypomagnesemia in this patient population. A prospective evaluation of the utility of sodium thiosulfate in preventing cisplatin-induced renal magnesium wasting appears indicated.