This study aims to compare 2 methods of assessing the postural phase of gait initiation, in regard to intrasession reliability, in healthy and poststroke subjects. As a secondary aim, this study aims to analyze anticipatory postural adjustments during gait initiation based on the center of pressure (CoP) displacements in poststroke participants. The CoP signal was acquired during gait initiation in 15 poststroke subjects and 23 healthy controls. Postural phase was identified through a baseline-based method and a maximal displacement-based method. In both healthy and poststroke participants, higher intraclass correlation coefficient and lower coefficient of variation values were obtained with the baseline-based method when compared with the maximal displacement-based method. Poststroke participants presented decreased CoP displacement backward and toward the first swing limb compared with controls when the baseline-based method was used. With the maximal displacement based method, there were differences between groups only regarding backward CoP displacement. Postural phase duration in medial-lateral direction was also increased in poststroke participants when using the maximal displacement based method. The findings obtained indicate that the baseline-based method is more reliable detecting the onset of gait initiation in both groups, while the maximal displacement-based method presents greater sensitivity for poststroke participants.