A model is proposed to account for cell survival after multiple doses, when the interfraction interval is insufficient for complete Elkind repair. In the limit of ever-increasing numbers of ever-smaller fractional doses, the model transforms into the accumulation model (Roesch, 1978) of survival after continuous irradiation. When it is adapted to describe tissue responses to isoeffective multifractionated regimens, wherein repair is incomplete, a generalization of the usually linear plot of reciprocal total dose versus dose per fraction is obtained, in which downward curvature is evident. There is some advantage in studying tissue responses to multifractionated regimens with incomplete repair in the interfraction intervals, or continuous exposures at various dose rates since in addition to determination of repair capacity (defined by beta/alpha) there is an estimate of repair kinetics (defined by the halftime T1/2 for repair of sublethal injury). There is a saving in overall treatment time with either method, thereby reducing the influence of regeneration on the interpretation of the results. The results of analyses of previously published data are presented to illustrate the use of the models. Estimated from the response of three acutely responding normal tissues in the mouse (jejunum, colon and bone marrow), repair halftimes ranged from 0.3-0.9 h and values of beta/alpha were approximately 0.1 Gy-1. From the response of mouse lung (LD50 for pneumonitis) to multifractionated regimens with incomplete repair, the repair halftime was estimated at 1.5 h and beta/alpha was 0.27 Gy-1. In the rat spinal cord beta/alpha was 0.7 Gy-1 and T 1/2 was 1.5 h.(ABSTRACT TRUNCATED AT 250 WORDS)