This study tested whether the potential doubling time of tumour cells measured before or during treatment could predict the repopulation rate of surviving clonogens during fractionated radiotherapy. Tumours used for the study were a fibrosarcoma (SSK 2), an adenocarcinoma (AT 7) and a squamous cell carcinoma (AT 478), all grown subcutaneously in the C3H mouse. Potential doubling times (Tpot) were measured using the thymidine nanalogue iododexyuridine (IUdR) and flow cytometry. Results were compared with previous radiobiological studies on these tumours in which repopulation rates during radiotherapy were estimated using the tumour growth delay and tumour cure assays. Fractionated treatments consisted of daily doses of 4 or 8 Gy to clamped (hypoxic) tumours, 6 days per week for 1-3 weeks. Tpot values increased markedly during therapy for two of the tumours (SSK 2 and AT 478), by a factor of more than 10 for AT 478 in the third treatment week. Tpot remained approximately constant for the third tumour (AT 7). In no case was there evidence from the labelling studies of a shortening of Tpot which would suggest accelerated repopulation. From the radiobiological data, effective clonogen doubling times during radiotherapy were calculated from the doses required to produce a given effect in short and long treatment schedules. In the second week of treatment, effective clonogen doubling times in two tumours were approximately equal to the pretreatment Tpot, and shorter than the pretreatment Tpot in the third tumour. At some time during treatment, the surviving clonogens in these tumours therefore proliferated at the same rate or faster than before treatment. The difference between the labelling and radiobiological measurements was ascribed to the fact that, shortly after the start of a fractionated treatment, the IUdR labelling technique measures primarily doomed cells. These results show that kinetic measurements using DNA labelling techniques made during fractionated radiotherapy in most cases do not reflect the proliferation status of the surviving cells which are responsible for treatment outcome. Pretreatment Tpot measurements give a much better indication of the proliferation rate of surviving cells but in some cases may underestimate repopulation during radiotherapy.