A retrospective evaluation of prognostic factors in 55 patients suffering from metastatic gestational trophoblastic disease (MGTD) treated by modified Bagshawe's CHAMOCA regimen was done. The prognostic significance of the eight prognostic factors in the WHO scoring system, number of sites of metastasis and FIGO staging were evaluated by univariate analysis using Chi-square test with Yates' correction and odds ratio and by multivariate analysis using Cox proportional hazard analysis and logistic regression analysis. In the univariate analysis, the intervals between antecedent pregnancy and the diagnosis of GTD, (P= 0.004) the level of hCG (P= 0.02) and the number of metastatic sites (P= 0.046) were significantly associated with death. In the multivariate analysis, only the interval between the antecedent pregnancy and the diagnosis and the level of hCG were significantly associated with death. Thus, it seems that the interval between antecedant pregnancy and the diagnosis and the level of hCG were the two most significant factors in predicting mortality in high risk MGTD. The WHO staging was more predictive of poor outcome than that of the FIGO staging in this group of patients.