Although a number of HIV prevention programs have been implemented, such as mass media campaigns, high rates of unprotected and concurrent sexual partnerships, as well as low uptake HIV testing and limited HIV knowledge, persist in Tanzania. We examined the effect and predicting factors of HIV prevention communication among people living with HIV (PLH) exposed to NAMWEZA intervention, and their at-risk social network members (NMs) Quantitative data were collected from 326 participants at baseline and 24 months of follow-up. In-depth interviews with 20 PLH were conducted at follow-up. Results indicated specific communication about condom use and HIV testing increased; (mean increase of 0.28 (SD = 0.14) scores, P = 0.012 and 0.42 (SD = 0.11) scores, p < 0.001 respectively while general discussion about protecting other people from HIV did not change significantly; mean increase was 0.01 scores (SD = 0.005), p = 0.890. Positive predictors of communication included being single; OR = 1.10, p = 0.01, female; OR = 1.15, p = 0.03, aged 30 years or older; OR = 1.23, p < 0.01, HIV knowledge, dose of NAMWEZA participation; OR = 1.01, p < 0.001, and high self-efficacy for condom use; OR = 1.4, p < 0.001. Stigma demonstrated a significant but negative association with communication for condom use; OR = 1.01, p < 0.01.Qualitative data reflected perceived possession of more individual skills and ability to address some personal/cultural obstacles to communicating about HIV prevention including those observed in the quantitative data. NAMWEZA improved communication about HIV prevention among PLH with their at-risk-NMs. The approach is a promising complement to media campaigns in similar populations. Future research and program evaluation efforts should explore how communities perceive and communicate about protecting others from HIV.