An in-vitro test was developed for predicting the efficacy of anti-tumour chemotherapy. Cell cultures were grown from freshly removed tumours and it was then demonstrated by DNA cytophotometry and immuno-cyto-chemistry whether the growing tumour cells corresponded to those of the original tumour cells. Several cytostatic agents were then tested for their efficacy of inhibiting growth at clinically customary dosage. Growing cell cultures were established in 306 of 413 submitted tumours (74%). They responded quite differently to the various drugs that were tried. The clinical course in 94 cases was observed for minimally four and a mean of eight months to obtain an in-vitro to in-vivo correlation of response, with 178 individual correlations. A discrepancy was recorded in 16% of cases, a false-positive in-vitro sensitivity result was 3.6 times more frequently associated with an in vivo resistance than the reverse. Concordance between test results and clinical tumour response occurred in 84%. The monolayer proliferation assay correctly indicated resistance in 93.8%, sensitivity in 72.8%.