Spontaneous and induced transmission of oncogenic retroviruses from parents to offspring is well documented in mammals, as exemplified by murine leukaemia viruses. In addition, it is now common knowledge that transgenic mice, carrying viral and cellular transforming genes, can show unusual tumour incidences and can transmit this character to successive generations. The present report deals with the occurrence of tumours in rodents exposed directly in utero to murine sarcoma viruses (MSV). Attention was focused on whether MSV are involved in the etiopathogenetic processes leading to early-life and adult-type neoplasms in mice. According to our data, this is a likely possibility, since, for example, Kirsten MSV (Ki-MSV) induced stage-specific tumours in CD-1 mice, including lung tumours and skin papillomas, following intraembryonal injections on gestation days 8 and 10, respectively. When the injections were given in the second half of pregnancy, however, only capillary angiomas and other vascular malformations of the brain, as well as mesenchymal sarcomas, were induced in the newborn. Harvey MSV (Ha-MSV), a close relative of Ki-MSV, and Moloney MSV (Mo-MSV) were much less dependent on the stage of embryogenesis, and mesenchymal sarcomas were frequently detected in prenatally infected animals. Other MSV viruses, including 3611, J-2 and 4070A, were active carcinogens in a way totally independent of the embryonal stage exposed, and tumours in the progeny ranged from mesenchymal sarcomas to lymphoblastomas and rhabdomyosarcomas, respectively. We next looked at the possibility that spontaneous tumours of CD-1 mice might result from the interaction of certain chemical carcinogens with developmentally expressed components of the Ki-MSV virus. The experiments conducted thus far have provided dubious results. An example is given by experiments with N-ethyl-N-nitrosourea (ENU). This compound is a powerful transplacental carcinogen in CD-1 mice, where it induces lung tumours. It was expected that by exposing developing mice to Ki-MSV on day 8 of embryogenesis and to ENU late in pregnancy or in newborns, a great increase in the incidence of lung tumours would have appeared in the offspring. However, this did not happen, and the agents behaved as if different cellular targets were involved. Understanding this difference could be very informative and useful in identifying neoplastic factors in mammalian embryos.