Abstract Background Prospective studies exploring the effects of psychosocial work factors on self-reported health (SRH) are lacking, especially those studying effect modifications. The objectives were to examine the prospective associations of these factors, and multiple exposures to these factors, with SRH in a national representative sample, and effect modifications by gender, age, and occupation. Methods The prospective study relied on the three data collection waves (2013, 2016, and 2019) of the national French Working Conditions survey and was based on a sample of 15,971 employees, in good SRH at the beginning of the follow-up period. The occupational exposures were time-varying variables measured in 2013 and 2016, and included: 20 psychosocial work factors grouped into 5 broad domains, 4 exposures related to working time/hours and 4 physical-biomechanical-chemical exposures. The incidence of poor SRH three years later was the outcome. Discrete time Poisson regression models were performed using weighted data and with adjustment for gender, age, marital status, life events, and occupation. Results Almost all the studied psychosocial work factors were predictive of poor SRH. Some physical-biomechanical-chemical exposures were found to predict poor SRH. Only rare effect modifications were observed according to gender, age, and occupation. Dose-response associations between multiple exposures and the incidence of poor SRH were observed for 4 among 5 domains of psychosocial work factors. Conclusions Our study underlined the effects of psychosocial work factors, as well as multiple exposure effects, on the incidence of poor SRH. However, most of these effects were the same across population groups related to gender, age, and occupation.