A simple colorimetric test, the MTT assay, has been adapted for chemosensitivity testing of human small cell lung cancer cell lines, and fresh tumour samples. Optimal conditions for clinical chemosensitivity testing were determined using established SCLC lines. Nineteen different chemotherapeutic agents were tested, and sixteen of them were found to be cytotoxic in this assay system. The drug sensitivity of a panel of 16 SCLC cell lines was measured and compared. There was very little intraexperiment variation, but the interexperiment variation was significant. Cell lines which were derived from patients who had not received chemotherapy at the time the cell line was established were more sensitive (to all but one of the drugs) than lines derived from treated patients, and the differences were statistically significant for two of the drugs. One cell line, NCI-H209, which was derived from an untreated patient, stood out as being the most sensitive or among the most sensitive to all of the drugs tested. Another cell line, H69AR, which is a multidrug resistant subline of the cell line NCI-H69, was the most resistant to many of the natural product drugs tested. Multiple drug chemosensitivity testing was performed on eight fresh tumour samples from SCLC patients (five pleural effusions, one lymph node, and two primary tumours). It was possible to perform chemosensitivity testing on all of the clinical samples in which sufficient tumour cells were available. The drug sensitivity of the clinical samples was, in most cases, within the same range as for the cell lines. Since this assay is very rapid and simple to perform, it may have practical applications in clinical drug sensitivity testing of human tumours.