This study examines whether working with a broker increases or reduces the payment received for the last client among female sex workers. Building on research on the informal economy and sex work, we formulate a positive embeddedness hypothesis, expecting a positive association, and an exploitation hypothesis, expecting a negative association. We analyze a large survey combined with intensive interview data on female sex workers in Andhra Pradesh, India. These data uniquely distinguish between the amount the sex worker actually received and the amount the client paid. The analyses show that brokers are associated with significantly lower last payment received. Although brokers are associated with a greater number of clients in the past week, this does not result in significantly higher total earnings in the past week. Further analyses suggest that much of the negative relationship with earnings is due to the fact that brokers lead to a lack of control over the amount clients are charged. At the same time, the results fail to show that brokers actually provide services of value. Ultimately, the results support the exploitation hypothesis. We conclude by encouraging the refinement of theories of embeddedness and exploitation and calling for greater research on workers in the informal economy of developing countries.