Idiopathic normal pressure hydrocephalus (iNPH) is a syndrome characterized by a triad composed of cognitive alteration, urinary incontinence, and gait impairment associated with ventricular enlargement and normal cerebrospinal fluid pressure. Gait impairment is among the earliest symptoms; however, the reliability of the evaluation is not well-established and no consensus has been reaching regarding variables that should be analyzed and which parameters should be considered to accurately assess post-intervention improvement. Are the degree of repeatability, standard error of measurement, and minimum detectable change considered to detect changes in gait variables in iNPH patients? A total of 84 iNPH patients with a mean age of 77.1 (±6.4) years were analyzed. Gait deviation index (GDI), speed, cadence, cycle time, stride length, single support, and first and second double support were chosen as the variables to be analyzed. Statistical analysis was performed by an independent evaluator, with gait repeatability assessed by the intraclass correlation coefficient (ICC) and the standard error of measure (SEM). ICC values were 0.76-0.85 with excellent repeatability, while SEM demonstrated that the variables with best repeatability were the GDI (mean, 4.94; 95% confidence interval (CI), 4.63-5.43), representing a 7.65% mean relative error of the measurement (mean, 0.05 m; 95% CI, 0.05-0.06), and stride length (mean 0.05 m; 95% CI, 0.05-0.06), with a 7.69% mean relative error. We concluded that GDI and stride length were the variables with the best repeatability and lower variability in the gait of iNPH patients.