Fifty Taiwanese patients with relapsed and/or refractory multiple myeloma (MM) were treated with thalidomide on a dose-escalation schedule, commencing with 100 mg/d nightly and incremented either to the maximally tolerated dose or 800 mg/d. Twenty-two patients (44%) responded, with 10 (45.5%) classified as partial remission and 12 (54.5%) minimal response (MR). Complete response did not occur. Of the 28 non-responders, 14 were progressive disease and 14 stable. The median time from commencement of thalidomide treatment to initial achievement of MR was 29 days (range, 8~155), and the corresponding thalidomide dose was 200 mg/d (range, 100~500). The median tolerated dose of thalidomide for the entire sample was 400 mg/d (range, 100~800), with only two (4%) able to tolerate 800 mg/d. Comparing responsive and non-responsive patients, statistically significant differences were not demonstrated for any characteristics except for CRP level and percentage cytogenetic change, which was slightly higher in the latter group relative to the former. Of particular interest, 18 of the 22 responders experienced transient reduction of leukocyte count preceding the attainment of significant reduction in M-proteins in comparison to only four of the 28 non-responders (82% vs. 14%; p<0.001). The median time from commencement of thalidomide treatment to attainment of minimal leukocyte count was 28 days (range, 7~150), with a mean of 2.19x10(9)/l (range, 0.96~3.35x10(9)/l). Leukopenia was generally transient, with rapid recovery despite subsequent continuation of thalidomide. Levels of other non-hematologically adverse effects attributed solely to thalidomide were generally acceptable. For 25 patients, thalidomide treatment was supplemented with low-dose dexamethasone (4 mg, every other day). Of these, 11 had relapsed from and 14 were primarily refractory to thalidomide treatment. Nine of the 25 dexamethasone-supplemented patients were responders (36%). Of particular note were the unusual events noted with this thalidomide-dexamethasone combination, including vascular thrombosis, acute cholecystitis, idiopathic interstitial lung disease and sudden cardiac death. Our results suggest that thalidomide is also effective for Taiwanese patients with refractory and/or relapsed MM. Importantly, the transient reduction in leukocyte count after commencement of thalidomide treatment may serve as a clinical predictor for response. Adverse effects should be carefully monitored when combining thalidomide and dexamethasone, however.