To assess DNA immunization as a strategy for protecting against HIV infection in humans, we utilized SIVmne infection of Macaca fascicularis as a vaccine challenge model with moderate pathogenic potential. We compared the efficacy of DNA immunization alone and in combination with subunit protein boosts. All of the structural and regulatory genes of SIVmne clone 8 were cloned into mammalian expression vectors under the control of the CMV IE-1 promoter. Eight M. fascicularis were immunized twice with 3 mg of plasmid DNA divided between two sites; intramuscular and intradermal. Four primed macaques received a further two DNA immunizations at weeks 16–36, while the second group of four were boosted with 250 μg recombinant gp160 plus 250 μg recombinant Gag–Pol particles formulated in MF-59 adjuvant. Half of the controls received four immunizations of vector DNA; half received two vector DNA and two adjuvant immunizations. As expected, humoral immune responses were stronger in the macaques receiving subunit boosts, but responses were sustained in both groups. Significant neutralizing antibody titers to SIVmne were detected in one of the subunit-boosted animals and in none of the DNA-only animals prior to challenge. T-cell proliferative responses to gp160 and to Gag were detected in all immunized animals after three immunizations, and these responses increased after four immunizations. Cytokine profiles in PHA-stimulated PBMC taken on the day of challenge showed trends toward Th1 responses in 2/4 macaques in the DNA vaccinated group and in 1/4 of the DNA plus subunit vaccinated macaques; Th2 responses in 3/4 DNA plus subunit-immunized macaques; and Th0 responses in 4/4 controls. In bulk CTL culture, SIV specific lysis was low or undetectable, even after four immunizations. However, stable SIV Gag–Pol- and env-specific T-cell clones (CD3 +CD8 +) were isolated after only two DNA immunizations, and Gag–Pol- and Nef-specific CTL lines were isolated on the day of challenge. All animals were challenged at week 38 with SIVmne uncloned stock by the intrarectal route. Based on antibody anamnestic responses (western, ELISA, and neutralizing antibodies) and virus detection methods (co-culture of PBMC and LNMC, nested set PCR- of DNA from PBMC and LNMC, and plasma QC-PCR), there were major differences between the groups in the challenge outcome. Surprisingly, sustained low virus loads were observed only in the DNA group, suggesting that four immunizations with DNA only elicited more effective immune responses than two DNA primes combined with two protein boosts. Multigenic DNA vaccines such as these, bearing all structural and regulatory genes, show significant promise and may be a safe alternative to live-attenuated vaccines.