Background: The available evidence suggests that people with epilepsy have reduced cardio-respiratory fitness and muscle strength endurance, and impaired balance and mobility. The 6-minute walk test (6MWT), Timed Up and Go (TUG), and 30-second chair stand test (30CST) are physical performance tests frequently used in clinical practice. Objective: To establish the test-retest reliability and minimal detectable change of the 6MWT, TUG, and 30CST in people with epilepsy. Methods: The study was designed as an observational study. Forty-one people with epilepsy (23 females, 18 males; mean age 34.7 ± 10.4 years) participated. The 6MWT, TUG, and 30CST were tested by a trained physiotherapist during two sessions, which were conducted 7–14 days apart. The test-retest reliability of measures was assessed using the intra-class correlation coefficients (ICC) using two-way random effects and absolute agreement methods. The 95% limits of agreement, standard error of measurement (SEM), and minimal detectable change (MDC₉₅) were calculated. Results: The 6MWT (ICC = 0.92, SEM = 15.8, MDC₉₅ = 43.8), TUG (ICC = 0.95, SEM = 3.2, MDC₉₅ = 0.5) and 30CST (ICC = 0.92, SEM = 1.0, MDC₉₅ = 2.8) performance measurements demonstrated excellent test-retest reliability. The 95% limits of agreement was calculated, as illustrated in a Bland-Altman plot. Conclusion: The 6MWT, TUG, and 30CST are reliable for measuring physical performance. The findings of this study can support researchers and clinicians to decide if a change score of a person with epilepsy is likely to be measurement error or true change.