To systematically review (1) psychometric properties of criterion isokinetic dynamometry testing of muscle strength in persons with poststroke hemiplegia (PPSH); and (2) literature that compares muscle strength in patients poststroke with that in healthy controls assessed by criterion isokinetic dynamometry.A systematic literature search of 7 databases was performed.Included studies (1) enrolled participants with definite poststroke hemiplegia according to defined criteria; (2) assessed muscle strength or power by criterion isokinetic dynamometry; (3) had undergone peer review; and (4) were available in English or Danish.The psychometric properties of isokinetic dynamometry were reviewed with respect to reliability, validity, and responsiveness. Furthermore, comparisons of strength between paretic, nonparetic, and comparable healthy muscles were reviewed.Twenty studies covering 316 PPSH were included. High intraclass correlation coefficient (ICC) inter- and intrasession reliability was reported for isokinetic dynamometry, which was independent of the tested muscle group, contraction mode, and contraction velocity. Slightly higher ICC values were found for the nonparetic extremity. Standard error of the mean (SEM) values showed that a change of 7% to 20% was required for a real group change to take place for most muscle groups, with the knee extensors showing the smallest SEM% values. The muscle strength of paretic muscles showed deficits when compared with both healthy and nonparetic muscles, independent of muscle group, contraction mode, and contraction velocity. Nonparetic muscles only showed minor strength impairments when compared with healthy muscles.Criterion isokinetic dynamometry is a reliable test in persons with stroke, generally showing marked reductions in muscle strength of paretic and, to a lesser degree, nonparetic muscles when compared with healthy controls, independent of muscle group, contraction mode, and contraction velocity.