In July 2020, we surveyed 604 female sex workers (Senegal_wave3_clean) and subsequently randomised them in December 2021 into two groups: a treatment group, which was offered pre-exposure prophylaxis (PrEP), and a control group, which was not offered PrEP (randomisation_clean). In April–May 2022, we collected follow-up data from 409 FSWs to study the effect of PrEP availability on risky sexual behaviours (Senegal_wave4_clean).The randomisation into treatment and control groups was stratified according to two criteria:Prior use of PrEP: Respondents were classified as having prior PrEP use if they had reported, during earlier survey waves in 2017 or 2020, participation in the 2015–2016 PrEP demonstration project in Senegal.Self-reported risk preferences in sex: In the July 2020 survey, respondents rated their sexual risk preferences on a Likert-like scale. Based on these responses, individuals were categorised into high-risk or low-risk groups.After generating the treatment and control lists, we accounted for potential recruitment caps imposed by the partner organisation responsible for rolling out PrEP. To maintain comparability, the treatment list was ordered such that the proportion of participants within each stratification category matched that of the control group as closely as possible. This approach ensures that, even if recruitment were to stop prematurely, the balance of stratified characteristics between those contacted in the treatment group and the control group would be preserved. If recruitment proceeded without limitations, following this specific order would not be necessary.