Mouse kidneys were bilaterally irradiated with fractionated doses of 15 X 1.2 Gy, 15 X 1.6 Gy or 15 X 2.0 Gy. After an interval of 26 weeks the mice were re-irradiated with a range of single doses and functional kidney damage was measured (from clearance of [51Cr]EDTA) at monthly intervals until 36 weeks after re-irradiation. Re-irradiation tolerance was assessed from a comparison of dose-response curves for renal damage in re-treated mice compared with age-matched controls, which received only the re-treatment single doses. Fractionated irradiation markedly reduced the tolerance to subsequent re-irradiation after 6 months. The data indicated a continuous progression of the initial damage, even when the doses used were below the threshold level required to give measurable functional damage. These results were qualitatively similar to those previously published for re-irradiation of mouse kidneys after single doses. A quantitative comparison of the re-treatment tolerance after fractionated and single-dose irradiation indicated, however, that fractionated irradiation to partial tolerance was less damaging than predicted on the basis of the linear-quadratic (LQ) model. These results suggest either the failure of the LQ model for renal irradiation damage at low doses per fraction or, possibly, the presence of two subpopulations in the kidney: one population responsible for overt functional damage and the second population leading to subthreshold damage not normally expressed as functional impairment.