The Posturomed device is used as a scientific tool to quantify human dynamic balance ability due to unexpected perturbations, and as a training device. Consequently, the question arises whether such measurements are compromised by learning effects. Therefore, this study aimed to analyze inter- and intra-day reliability of dynamic balance responses using the Posturomed. Thirty healthy young subjects participated (24.3±3.2 years). The Posturomed was equipped with a triggering mechanism to enable unexpected, horizontal platform perturbations. A force platform was used to quantify Center of Pressure (COP) excursions for two time intervals: interval 1 (0–70 ms post perturbation) and interval 2 (71–260 ms post perturbation). Dynamic balance tests were performed in single leg stances in medio-lateral and anterior-posterior perturbation directions. Inter- and intra-day reliability were assessed descriptively using Bland-Altman plots and inferentially using tests for systematic error and intra-class-correlations. With regard to the mean COP excursions for every subject and all intervals, some cases revealed significant differences between measurement sessions, however, none were considered relevant. Furthermore, intra class correlation coefficients reflected high magnitudes, which leads to the assumption of good relative reliability. However, analyzing inter- and intra-day reliability using Bland-Altman plots revealed one exception: intra-day comparisons for the anterior-posterior direction in interval 2, which points towards possible learning effects. In summary, results reflected good overall reliability with the exception of certain intra-day comparisons in the anterior-posterior perturbation direction, which could indicate learning effects in those particular conditions.