A recently developed method for explicitly evaluating the time-averaged performance measure of an impulse-forced dynamical system is used to evaluate the effect of periodic injections as a cancer chemotherapy regimen. Explicit formulae are derived for two time-averaged performance measures, thus avoiding the need for tedious repeated numerical integrations. The method is shown to give more accurate predictions than previous periodic optimization methods in the case of one performance measure, and is successfully applied to the other performance measure for which previous optimization methods are inapplicable.