We previously showed that a carrageenan (CG) gel containing 50 μM MIV-150 (MIV-150/CG) reduced vaginal simian/human immunodeficiency virus (SHIV)-RT infection of macaques (56%, p>0.05) when administered daily for 2 weeks with the last dose given 8 h before challenge. Additionally, when 100 mg of MIV-150 was loaded into an intravaginal ring (IVR) inserted 24 h before challenge and removed 2 weeks after challenge, >80% protection was observed (p<0.03). MIV-160 is a related NNRTI with a similar IC(50), greater aqueous solubility, and a shorter synthesis. To objectively compare MIV-160 with MIV-150, herein we evaluated the antiviral effects of unformulated MIV-160 in vitro as well as the in vivo protection afforded by MIV-160 delivered in CG (MIV-160/CG gel) and in an IVR under regimens used with MIV-150 in earlier studies. Like MIV-150, MIV-160 exhibited potent antiviral activity against SHIV-RT in macaque vaginal explants. However, formulated MIV-160 exhibited divergent effects in vivo. The MIV-160/CG gel offered no protection compared to CG alone, whereas the MIV-160 IVRs protected significantly. Importantly, the results of in vitro release studies of the MIV-160/CG gel and the MIV-160 IVR suggested that in vivo efficacy paralleled the amount of MIV-160 released in vitro. Hundreds of micrograms of MIV-160 were released daily from IVRs while undetectable amounts of MIV-160 were released from the CG gel. Our findings highlight the importance of testing different modalities of microbicide delivery to identify the optimal formulation for efficacy in vivo.