OBJECTIVE: To evaluate the reliability and variability of repeated measurements of dynamic and static lumbar muscle endurance. DESIGN AND SETTING: Participants performed an isometric lumbar-extension strength test followed by 2 trials of 4 separate lumbar muscular-endurance tests (with a 24-hour rest period between tests). Data were collected at a university musculoskeletal research laboratory. SUBJECTS: Eight healthy, physically active volunteers (5 men, 3 women; age = 25.9 +/- 4.3 years; height = 169.0 +/- 4.6 cm; mass = 73.9 +/- 33.1 kg) participated in this investigation. MEASUREMENTS: We initially tested each participant's isometric lumbar-extension peak torque on a lumbar-extension dynamometer. Static (holding time) and dynamic (repetitions) lumbar-endurance tests were subsequently performed on the lumbar-extension dynamometer and a horizontal roman chair. RESULTS: Interclass reliability was high for all endurance tests completed (r = 0.91 to 0.96, P </=.05). Variability (expressed as total error) for the static-dynamometer and roman-chair tests was 18.3 and 11.6 seconds, respectively, with 2.8 and 1.6 repetitions for the dynamic-dynamometer and roman-chair tests, respectively. CONCLUSIONS: Lumbar muscle endurance can be reliably assessed by both static and dynamic protocols on high- and low-technology devices.