We present results from a preliminary system dynamics model of problems in recruiting clients to a hypothetical HIV prevention program. Efforts in HIV prevention emphasize moving programs of demonstrated efficacy to community settings. However, little is known about how these programs interact with contextual elements of service delivery to determine the feasibility of implementation. The section of the model we present here focuses on the stocks and flows associated with attracting, enrolling, and graduating a steady flow of clients into small-group workshops and highlights paradoxes in providing this type of program in the community. We test two policies that either focus on monitoring the recruitment rate or monitoring the graduation rate. Despite its superiority in real-life experiments for producing behavioral change, our model suggests that small-group workshops are a highly inefficient means to change the behavior of a target population over a ten-year period of time.