Markov models are tools used in decision analyses. Named after a 19th-century Russian mathematician, Markov models are the basis of software programs that model what might happen to a cohort of patients during a series of cycles (eg, periods of 1 year). The model allows for the possibility that patients might move from one health state to another. For instance, one patient may have a mild stroke in one 3-month cycle, continue with minimal functional limitation for a number of cycles, have a gastrointestinal bleeding episode in a subsequent cycle, and finally experience a major stroke. Ideally, data from randomized trials will determine the probability of moving from one state to another during any cycle under competing management options.