BACKGROUND: Improvement of balance control is an important rehabilitation goal for patients with motor and sensory impairments. To quantify balance control during walking, various stability outcome measures have described differences between healthy controls and patient groups with balance problems. To be useful for the evaluation of interventions or monitoring of individual patients, stability outcome measures need to be reliable. RESEARCH QUESTION: What is the test-retest reliability of six stability outcome measures during gait? METHODS: Patients with balance problems (n = 45) and healthy controls (n = 20) performed two times a two-minute walk test (2MWT). The intraclass correlation coefficient (ICC) and Bland-Altman analysis (coefficient of repeatability; CR) were used to evaluate the test-retest reliability of six stability outcome measures: dynamic stability margin (DSM), margin of stability (MoS), distance between the extrapolated centre of mass (XCoM) and centre of pressure (CoP) in anterior-posterior (XCoM-CoPAP) and medial-lateral (XCoM-CoPML) direction, and inclination angle between centre of mass (CoM) and CoP in anterior-posterior (CoM-CoPAP-angle) and medial-lateral (CoM-CoPML-angle) direction. A two way mixed ANOVA was performed to reveal measurement- and group-effects. RESULTS: The ICCs of all stability outcome measures ranged between 0.51 and 0.97. Significant differences between the measurements were found for the DSM (p = 0.017), XCoM-CoPAP (p = 0.008) and CoM-CoPAP-angle (p = 0.001). Significant differences between controls and patients were found for all stability outcome measures (p < 0.01) except for the MoS (p = 0.32). For the XCoM-CoP distances and CoM-CoP angles, the CRs were smaller than the difference between patients and controls. SIGNIFICANCE: Based on the ICCs, the reliability of all stability outcome measures was moderate to excellent. Since the XCoM-CoPML and CoM-CoPML-angle showed no differences between the measurements and smaller CRs than the differences between patients and controls, the XCoM-CoPML and CoM-CoPML-angle seem the most promising stability outcome measures to evaluate interventions and monitor individual patients.