The influence of age at the time of irradiation on the lifetime risk for excess mortality from solid tumors, and on the temporal pattern of variation in the excess mortality rate, was analyzed using data obtained from a study of female B6C3F1 mice, which was conducted at the National Institute of Radiological Sciences, Chiba, Japan. Mice were irradiated with 1.9-Gy gamma rays at day 17 in intra-uterine age, or day 0, 7, 35, 105 or 365 in postnatal age. Control and irradiated mice were allowed to live out their entire life span under a specific pathogen-free condition. The primary cause of death for each mouse was determined by macroscopic and microscopic examination. The lifetime excess mortality from solid tumors was apparently higher in the mice irradiated during the neonatal to puberty period than in the mice irradiated during the intra-uterine or adult period. The median of time for manifestation of lifetime excess mortality since irradiation was shortest among mice exposed at 365 days of age and longest among mice exposed at 17 days of intra-uterine age. The excess mortality rate at any attained age was not independent of the age at irradiation. The excess mortality rate increased with increasing age, and the excess relative risk decreased with increasing age. The temporal variations of the excess mortality rate and background mortality rate were analyzed using the additive multi-stage model, which includes the assumptions that radiation-related carcinogenesis superimposes on background carcinogenesis, and that both radiation-related and background carcinogenesis involve multiple stages. The results of the analysis strongly suggested that the number of stages for manifestation of radiation-related carcinogenesis was less than that in background carcinogenesis for various types of solid tumors, and that the majority of stages were common in both radiation-related and background carcinogenesis. The additive multi-stage model well described the observed findings on the length of the latent period and temporal variations of the excess mortality rate and excess relative risk. It should be stressed that the magnitude of the lifetime risk was not only determined by a decrease in the number of hits for carcinogenesis but was also determined by another parameter which decides the initial value of excess mortality rate. Furthermore, we estimated the rate of decrease in the number of remaining hits for carcinogenesis, and it was found that the rate of decrease in the number of remaining hits was higher in several irradiated groups than that in the background carcinogenesis. However, radiation-induced genomic instability and/or delayed mutation may be of secondary importance when radiation was delivered promptly, because the present analysis revealed that the major action of radiation took place soon after irradiation, as one or more hits for transitions of stages for carcinogenesis.