Using data from a randomized controlled trial involving HIV education, home-based HIV counseling and testing (VCT), and conditional cash transfer (CCT) for facility-based VCT, we find home-based VCT and CCT to facility-based VCT to significantly increase HIV test take-up, with the former being more effective in detecting HIV-positive cases. While HIV education increases HIV knowledge in the short run, the HIV-knowledge gap between the treated and control groups dissipates over time which could be due to knowledge spillover. We find updating of infection expectation among testers and an increase in the demand for condom among HIV-positive testers, but both effects dissipate over time, in six months.