Public health approaches to interrupting infectious disease transmission have yet to be informed by traditional deterministic models of contagion. We investigate this gap in current violence prevention research by introducing a Susceptible–Transmitter–Victim Epidemic model, based on the classic Susceptible–Infectious–Recovered differential equation model, to explore the impact of violence interruption on the diffusion of violence. Uncertainty and sensitivity analysis are done using Latin hypercube sampling. Based on sensitivity analysis results, model predictions appear to be overestimating annual gun assault cases, where the mean estimate of the gun assault rate at equilibrium is double the average gun assault rate over the past decade. Several key parameters are identified as significant to gun assault predictions and may account for model imprecision. Scenario analysis is also done to determine the effectiveness of violence interruption programs. Results suggest that targeting all potential violence transmitters can reduce gun violence three times more than an intervention that only targets gun-owning individuals, indicating the importance of taking a holistic approach to violence interruption and prevention. Our results also suggest that having individuals in the population transmitting violence, whether or not they are participating in gun violence, is sufficient to sustain a gun violence epidemic.