A method for analyzing reaction times (RTs) in tasks involving both sequential and concurrent processing is proposed. Tasks are analyzed with the method by selectively prolonging mental processes, as with the additive factor method. Falsifiable predictions about the changes in RT produced by prolonging processes are derived by drawing on the theory of scheduling. Under certain conditions, which frequently arise in practice, one can determine for a given pair of processes whether they are executed sequentially or concurrently. If one process precedes another, one can often determine which comes first. One can also construct intervals within which the process durations lie. Two experiments are analyzed using the method. One, by Holyoak, Dumais, and Moyer, is on a sentence-verification task involving associated and unassociated items. The other experiment is on the Stroop effect and supports the single-channel hypothesis that a subject makes only one decision at a time. The data suggest that in the color-naming task the decision about the word precedes the decision about the color, whereas in the word-naming task the order of the decisions is reversed.