The actual possibilities to test chemical substances for carcinogenic activities are critically assessed. Short-term-tests are especially discussed. By these tests--now utilized for screening of chemicals--specific biological activities can be determined, which correlate with carcinogenic effects in vivo. The phase of checking and improving the vitro-tests is not yet finished with regard to the rate of correlation of the results from those tests and the results from long-term experiments in animals as well as from clinical and epidemiological human data. Parallel to this investigations, results from long-term-tests should be critically evaluated, because the experimental induction of tumors in animals is up to now the only way to prove carcinogenic activity of a chemical. Short-term-tests should be utilized as screening methods for the selection of suspicious substances, which should or must be further tested in long-term animal experiments. At present results from short-term-tests are not sufficient to classify a substance as a carcinogen or a non-carcinogen. Certain short-term-tests can and will get significance for the evaluation of the carcinogenic risk of chemicals to man when comparing the metabolism of carcinogens in vitro using enzymatic systems of a series of mammalians including man.