A decision algorithm was required to evaluate the first half of a cooperative agreement for preclinical trials to optimize medical device parameters to enhance stroke recovery. Continued funding was contingent upon the midpoint evaluation, called the milestone decision. We developed an algorithm, which summarized our rodent and primate model results. Primary outcomes weighed more heavily than secondary outcomes, and the primate model more heavily than rodent models. By controlling the type I error for this milestone decision, the probability of correctly discontinuing the study if treatment was not beneficial was high (>0.84). Similar algorithms may be adapted for other milestone-driven projects.