We have investigated the role of host immunological factors in the formation of "tumor colonies" in the spleens of unirradiated C57BL/6 X C3Hf/Bi FI mice 9 days after i.v. injection of spleen cells from Friend virus (FV)-infected C3Hf/Bi donors. Pretreatment of hosts with antilymphocyte serum (ATS) increased the number of tumor colonies. Pretreatment with formalinized FV-infected cells had the opposite effect, and ATS diminished the inhibitory effect of preimmunization. Cell suspensions from 11 individual FV-infected donors were examined. The suspensions differed with respect to their behavior on transplantation into untreated and ATS-pretreated F1 hybrid hosts. With several suspensions, the number of tumor colonies produced was approximately proportional to the number of cells injected; in all of these, ATS increased the slope of the line relating colony number to cell number. With most of the suspensions, tumor colony-forming efficiencies in untreated hosts strikingly decreased with increasing number of cells injected; ATS induced an increase in the number of tumor colonies and rendered the colony-forming response more nearly proportional to cell number. With two suspensions, few or no colonies developed; pretreatment with ATS had no significant effect. When the 11 cell suspensions were considered together, a proportional relation was found between the magnitude of the ATS effect (i.e., colony number in the presence of ATS minus colony number in the absence of ATS) and the colony-forming efficiency in ATS-treated mice. The ATS effect on the average was equivalent to a 2-fold increase in tumor colony-forming efficiency. We interpret these findings to indicate that two factors interact to determine the number of tumor colonies produced by spleen cells from FV-infected C3H donors in untreated F1 hybrid hosts. One is a property of the FV-infected cell population and includes its frequency of tumor colony-forming units; this factor varies widely among different cell suspensions. The other is a property of the tumor colony-forming units-host interrelationship and includes the vulnerability of tumor colony-forming units to the host immune response elicited by the injected cells; this factor appears to be constant with different cell suspensions. The present results show that the two factors can be dissociated in immunosuppressed hosts.