In situ hybridization has been used as a quantitative method to study the replication of SV40 during lytic infection. Various parameters have been defined such as fixation and denaturation of cytological preparations, concentration of complementary (3H)-cRNA and volume of buffer used for the hybridization reaction. When all these parameters are carefully controlled, a reproducibility of +/-10% can be obtained for the quantitative study of the system SV40-monkey kidney cells (CV-1). In this system, it is evident that the synthesis of viral DNA is not synchronized in confluent cells. Moreover, a proportion of 20 to 30% of cells are not labelled whatever the multiplicity of infection. In SV40-transformed cells, it was not possible to detect significantly the integrated genome due either to the small size of the genome or to the fact that the number of genome-equivalents integrated at the same place is too low to yield a radioactivitity superior to background level for long periods of exposure.