Paramecia cells, like human diploid cells cultured in vitro, provide a useful model system for understanding the mechanism that limits division potential. The reported maxima of the clonal lifespan of Paramecium tetraurelia fall into two ranges: from 220 to 258 fissions and from 310 to 325 fissions. We found that neither the selection of vigorous lines nor the cryptic occurrence of autogamy offers a plausible explanation for the much longer lifespans in the latter range. We found the sporadic occurrence of very long clonal lifespans, such as 330 fissions, without selection and autogamy. Selection, which was evaluated by using different methods to maintain lines, had little effect on the extension of the maximal clonal lifespan, whereas it did have a marked effect on the extension of the mean clonal lifespan. Autogamy, which was checked with two closely linked marker genes, was frequent, but only during the period when lines were terminating. Statistical analysis on the mean clonal lifespans of two groups of subclones cultured at 25 and 20 degrees C or with rich and poor nutrition showed that the clonal lifespan that was fission-related under favourable conditions tended to dissociate from fissions under less favourable conditions. We discuss mechanisms that determine the clonal lifespan, programmed events contributing to the maximum clonal lifespan and random events contributing to the mean clonal lifespan.