In many animal model systems for carcinogenesis, well characterized putative premalignant lesions are observed. A much studied example is provided by the enzyme altered foci in rodent hepatocarcinogenesis experiments. In a recent paper, we proposed a method for the quantitative analysis of such premalignant lesions. The model used in that paper assumed that the mean growth of premalignant clones is exponential. However, it has been suggested that such a model is oversimplified. In this paper, we relax the assumption of exponential mean growth. The new model contains one extra parameter that measures departures from exponentiality. Use of the model is illustrated by analysis of ATPase deficient foci in the liver of rats given NNM (N-nitrosomorpholine) in their drinking water. The analysis suggests that the clonal growth of altered cells is significantly accelerated (superexponential) for nontoxic doses of NNM. Finally, the hazard function of the two-mutation model for carcinogenesis is briefly discussed under nonexponential (mean) growth of intermediate cells.