Activation of prm, the promoter at which cI transcription is initiated in immune bacteriophage lambda lysogens, requires no phage genes or sites outside the cI region, cro product, however, is able to reduce fivefold the efficiency of cI expression from this promoter. cro product does not inactivate λ repressor, the cI gene product, and cro product also acts effectively in the presence of active repressor. Therefore, cro product seems to block directly the promoter prm at a site which can not be protected by repressor binding to the right operator Or. Since cro product remains an effective inhibitor of prm in a bacterial strain in which it can not block efficiently leftward transcription of the N-cIII operon or rightward transcription of the cro-cII operon, cro product probably has three separate target sites; one near p1 at which it regulates N and cIII, one near pr at which it regulates cro and cII, and a third near prm at which it blocks cI. My experiments indicate that repressor must be bound to the right operator Or to activate either the promoter prm or a post-transcriptional step in cI polypeptide synthesis. In the absence of cro product, cI expression from prm is not constitutive. Pre-existing repressor stimulates sevenfold the initial rate of active repressor synthesis by λ NcrocII. After 20 generations growth at high temperatures, only 3 to 10% the normal repressor antigen level is seen in defective, cro-minus lysogens containing any of four different cIts alleles. Studies on the cIts857 antigen, presented in the Appendix, show that these low antigen levels cannot simply be a consequence of poor antigen formation and stability. Instead they must reflect at least a fourfold decrease in the rate of cI polypeptide synthesis. The weakly constitutive right operator mutation v3 reduces the maximum temperature at which cIts857 repressor synthesis can be maintained in cro-minus lysogens. cro+ phage bearing more constitutive right operator mutations block repressor synthesis in trans by making cro product. These operator mutations are much more effective at blocking repressor synthesis in cis, however, which also argues that repressor must be bound to the operator if there is to be efficient cI polypeptide synthesis initiated at the promoter prm.