Exogenously added N6, 02'-dibutyryladenosine-3'-5'-cyclic phsophate (DBcAMP) exerted a biphasic effect on the in vitro primary response of mouse spleen cells to sheep erythrocytes (SRBC). High concentrations of DBcAMP were markedly suppressive, whereas lower concentrations brought about significant increases in the peak numbers of antibody-forming cells. The kinetics of the response were not altered by DBcAMP. A direct effect of DBcAMP on bone-marrow derived lymphocytes was suggested because this nucleotide increased the in vitro response of "nude" mouse spleen cells to SRBC. The magnitude of the response, however, was considerably less than that mounted by spleen cells of normal mice. These results show that DBcAMP can have either enhancing or suppressive effects on antibody formation and these effects are dependent upon nucleotide concentration, antigen-dosage and time of assay. The effects of this cyclic nucleotide differed in two major respects from those of the synthetic polyribonucleotide adjuvant, polyadenylic acid-polyuridylic acid (poly A:U) complexes. Thus poly (A:U) decreased the time of peak response in vitro but did not increase the magnitude of the response.