Objective: To assess the reliability of knee extensor and flexor muscle strength measurements in persons with late effects of polio. Design: A test-retest reliability study. Subjects: Thirty men and women (mean age 63 (standard deviation 6.4) years) with verified late effects of polio. Methods: Knee extensor and flexor muscle strength in both lower limbs were measured twice 7 days apart using a Biodex dynamometer (isokinetic concentric contractions at 60°/sec and isometric contractions with knee flexion angle 90º) and a Leg Extension/Curl Rehab exercise machine with pneumatic resistance (HUR) (isotonic contractions). Reliability was assessed with the intraclass correlation coefficient (ICC1,1), the mean difference between the test sessions (đ) together with the 95% confidence intervals for đ, the standard error of measurement (SEM and SEM%), the smallest real difference (SRD and SRD%) and Bland-Altman graphs. Results: Test-retest agreements were high, (ICC1,1 0.93-0.99) and measurement errors generally small. The SEM% was 4-14% and the SRD% 11-39%, with the highest values for the isokinetic measurements. Conclusion: Knee muscle strength can be measured reliably and can be used to detect real changes after an intervention for a group of persons with late effects of polio, whereas the values may be too high for single individuals or to detect smaller short-term changes over time for a group of individuals.