A variety of test systems have been developed for predicting the efficacy of cytotoxic drugs in the treatment of individual malignant human tumors. The present paper reports on the author's experience with Volm's short-term chemosensitivity test and Hamburger and Salmon's stem-cell assay. In the Volm test the influence of adriamycin on the incorporation of radioactively labelled uridine in an individual tumor cell suspension was investigated. Comparison with a cytostatic-free control permitted conclusions to be drawn with regard to the proliferation-related chemosensitivity. The stem-cell assay is based on the capacity of certain tumor cells, the so-called stem cells, to form colonies in a bilaminar soft agar system. The growth of the colony of pre-incubated cytostatics was evaluated in relation to that of untreated tumor cells. The Volm test was successful in 63 (95%) out of a total of 66 tests conducted. Twelve tumors were chemosensitive in the test and 51 chemoresistant. In the stem-cell assay, growth of a colony which permitted chemosensitivity to be tested was only found in 27 out of 183 tests. The criterion of chemosensitivity with a reduction of at least 50% in the number of cells in the colony by at least one cytostatic was satisfied by 14 (29%) of the 49 stem-cell assays which could be evaluated, there being no differences between breast and ovarian carcinomas. The two test systems indicated the chemosensitivity correctly in less than 50% of 49 retrospectively evaluated courses of disease. In contrast, resistance was predicted correctly in 90%.(ABSTRACT TRUNCATED AT 250 WORDS)