The estimates of a dose-response model for human skin cancer (non-melanotic) incidence have been compared with the observed records for the Bristol area over the period 1952--1972. In general the model appears to respond more slowly than the real population. The trend of the model expectations agrees qualitatively with the observed data and correlation of annual incidences, although negligible in the first decade, is very good in the second (corr. factor = 0.74). The results indicate that an increase of 1% in dose might eventually lead to an increase of between 2% in skin cancer incidence.