BACKGROUND: The incidence of HIV infection among female sex workers and their clients in the Middle East and north Africa is not well known. We aimed to assess HIV incidence, the contribution of heterosexual sex work networks to these numbers, and the effect of interventions by use of mathematical modelling. METHODS: In this modelling study, we developed a novel, individual-based model to simulate HIV epidemic dynamics in heterosexual sex work networks. We applied this model to 12 countries in the Middle East and north Africa that had sufficient data to estimate incidence in 2020 and the impact of interventions by 2030 (Algeria, Bahrain, Djibouti, Iran, Libya, Morocco, Pakistan, Somalia, South Sudan, Sudan, Tunisia, and Yemen). Model-input parameters were provided through a systematic review of HIV prevalence, sexual and injecting behaviours, and risk group size estimates of female sex workers and clients. Model output was number of incident HIV infections under different modelling scenarios for each country. Summary statistics were generated on these model output scenarios. FINDINGS: Based on the output of our model, we estimated a total of 14 604 (95% uncertainty interval [UI] CI 7929-31 819) new HIV infections in the 12 countries in 2020 among female sex workers, clients, and spouses, which constituted 28·1% of 51 995 total new cases in all adults in these 12 countries combined. Model-estimated number of new infections in 2020 in the 12 countries combined was 3471 (95% UI 1295-10 308) in female sex workers, 6416 (3144-14 223) in clients, and 4717 (3490-7288) in client spouses. Contribution of incidence in heterosexual sex work networks to total incidence varied widely, ranging from 3·3% in Pakistan to 71·8% in South Sudan and 72·7% in Djibouti. Incidence in heterosexual sex work networks was distributed roughly equally among female sex workers, clients, and client spouses. Estimated incidence rates among female sex workers per 1000 person-years ranged from 0·4 (95% UI 0·0-7·1) in Yemen to 34·3 (17·2-59·6) in South Sudan. In countries where HIV acquisition through injecting drug use creates substantial exposure for female sex workers who inject drugs, estimated incidence rates per 1000 person-years ranged from 5·1 (95% UI 0·0-35·1) in Iran to 45·8 (0·0-428·6) in Pakistan. The model output predicted that any of the programmed interventions would substantially reduce incidence. Even when a subpopulation did not benefit directly from an intervention, it benefited indirectly through reduction in onward transmission, and indirect impact was often half as large as the direct impact. INTERPRETATION: Substantial HIV incidence occurs in heterosexual sex work networks across the Middle East and north Africa with client spouses being heavily affected, in addition to female sex workers and clients. Rapid scaling-up of comprehensive treatment and prevention services for female sex workers is urgently needed. FUNDING: Qatar National Research Fund (a member of Qatar Foundation), the Biostatistics, Epidemiology, and Biomathematics Research Core at the Weill Cornell Medicine-Qatar, Qatar University-Marubeni, the UK Medical Research Council, and the UK Department for International Development.