p53 is over-expressed in approximately 50% of human cancers, and transfer of cytotoxic T lymphocytes (CTL) against wild-type p53 protects mice against p53-over-expressing tumors, suggesting that p53 might be an attractive target for immunotherapy. Immunization of mice with a recombinant canarypox virus, ALVAC, expressing human wild-type p53 (vCP207) prevented growth of p53-over-expressing tumors. Since intravenous administration induced better immune responses in mice than other routes, we have proposed to use this route in cancer patients. However, because this vector has never been administered intravenously to humans, and because of the possibility of inducing auto-immunity to a self-antigen, we felt it was necessary to first evaluate safety in rhesus macaques. We found that three intravenous administrations of vCP207 at proportional doses up to 10x those proposed for humans produced no abnormalities in hematologic or clinical chemistry parameters. Serologic markers of autoimmunity and inflammation were unaffected, despite the >95% amino acid identity between human and rhesus p53. Pathological examination of numerous tissues yielded findings comparable to those in animals given placebo. Some animals showed anti-p53 antibody responses following vaccination, indicating that tolerance could be broken to some extent. However, with the exception of one animal with a possible delayed type hypersensitivity reaction to p53 protein, we did not see evidence for a cell-mediated response. The safety profile in monkeys with ALVAC-p53 provides encouragement for using such live, modified vectors via the intravenous route for human immunotherapy.