We investigate the test-retest reliability and construct validity of gait measured with the DynaPort MiniMod under single and dual task conditions in diabetic patients walking on irregular terrain to identify the measurement error (precision) and minimal clinical detectable change.