Synthetic cytosine-phosphorothioate-guanine (CpG) oligodeoxynucleotides mimic immunostimulatory bacterial DNA and bind to TLR9 present on APCs, resulting in their activation. CpG have potent anti-tumor effects and are in clinical trials. The systemic effects of CpG given in the peri-BMT period are unknown. We now report that GVHD was markedly increased by weekly B-class CpG administration in 2 distinct full MHC disparate systems. As compared to controls, CpG reproducibly resulted in a significant decrease in survival rates (37–50% vs 0%). CpG induced proinflammatory cytokine release (IL12p70; IFNγ; TNFα; IL6), indicative of APC activation, as measured on post-BMT days 2, 4, 6, and 8 in recipients of BM or BM and spleen. On day +14, in vivo imaging using donor GFP transgenic T cells revealed a profound increase in T cell infiltration in secondary lymphoid organs (spleen, Peyers patches) and in other GVHD target organs (liver, lung, skin, ileum, colon), indicating that accelerated lethality was not due exclusively to cytokine-induced tissue injury. To determine whether donor or host APCs contributed to CpG-accelerated GVHD, CpG was given to wild type (wt) or TLR9 deficient recipients of allogeneic wt or TLR9 deficient donor splenocytes. Host but not donor TLR9-expressing cells were essential for GVHD acceleration. To determine which cytokine/receptor interactions were required for CpG-induced GVHD, studies were performed with cytokine/receptor deficient hosts. Either IFNγ or TNF, but not IL-6 or IL-12p40, was required for increased GVHD. Because CpG augmented donor anti-host allogeneic responses, we studied the effects of CpG on host anti-donor mediated BM rejection. CpG given on days 0 and 7 to sublethally irradiated (6.0–6.25 Gray) recipients of full MHC disparate increased host rejection of T cell depleted (TCD) BM, resulting in aplasia seen in 50–100% of CpG-treated vs 0–10% of controls. In sublethally irradiated (5.5 Gray) recipients of single MHC disparate TCD BM, mean donor chimerism levels in peripheral blood significantly decreased from 91% to 37% (MHC class I system) and 77% to 39% (MHC class II system) for control vs CpG-treated mice, indicating CpG stimulated host CD8− and CD4− mediated rejection. A brief incubation of TCD BM with CpG prior to BMT was not sufficient to cause rejection. CpG-treated recipients deficient in IL6, IL12p40, IFNγ or TNFR1 all had increased rejection, indicating that none of these cytokines were essential for promoting graft rejection. To determine whether donor or host TLR9 ligation stimulated graft rejection, CpG were given to wt or TLR9 deficient recipients of allogeneic wt or TLR9 deficient, TCD BM. In contrast to GVHD, donor and not host TLR9 ligation was essential for stimulating BM rejection. These data indicate a role of donor APCs in allogenenic BM rejection and provide further insight into the role of TLR9 signaling in GVHD. Our studies have important implications for the testing of CpG in clinical trials.