The SOAR model is used to examine several issues surrounding the interdiction of cocaine shipments into the United States. The model suggests that back-of-the-envelope estimates of drug interdiction's effectiveness may be overly optimistic if they neglect the existence of a backstop technology (smuggling small shipments over land), the concavity of smuggler costs as a function of the fraction of all routes on which the interdiction rate is increased, and the fact that not all smuggling costs are caused by interdiction. Considering these factors, it appears that one would expect increasing interdiction to have a substantial impact on U.S. cocaine consumption only under truly exceptional circumstances. The authors conclude that SOAR represents an initial effort to systematically analyze how drug interdiction can raise smuggler costs and lower drug consumption. 14 references, 4 tables, and 6 figures