BACKGROUND: To reach 2030 targets, the World Health Organization specifically recommends test-and-treat strategies to eliminate HIV and bacterial STI. Mathematical models identified the required coverages for the success of these interventions. However, they appear out of reach for current testing protocols based on medical recommendations. Still, promoting voluntary testing can effectively increase testing rates if individuals perceive high benefits associated to frequent testing. Design of multi-disease voluntary testing kits remains a key strategic tool to reach various epidemiological goals. METHODS: We construct mathematical models to describe individual-level decision-making about testing within a given epidemiological context. In particular, we discuss the four-disease epidemic of HIV, syphilis, chlamydia and gonorrhoeae in the Parisian community of men-who-have-sex-with-men. We acknowledge that individuals undergo pre-exposure prophylaxis against HIV that requires periodic testing. Furthermore, individuals may be biased in estimating their HIV risk relative to the risk of acquiring bacterial STI. Under these circumstances, we identify the combined testing strategies which bring HIV closest to elimination. RESULTS: An unbiased perception of disease risk works best toward HIV elimination. Secondly, regardless PrEP coverage and perception bias, combining HIV and syphilis voluntary testing performs worse than testing for HIV alone or along with other STI such as chlamydia and/or gonorrhoeae, to achieve HIV elimination. CONCLUSIONS: We argue against combining testing for HIV and syphilis as a voluntary testing strategy. Rather, to bring HIV closer to the elimination, it is best that voluntary testing for HIV and syphilis requires two distinct individual-level decisions.