Abstract Background The current study investigated the repeatability (test-retest reliability) of ground reaction force parameters recorded during a voluntary step execution under single (motor task) and dual task (motor and cognitive task) conditions for healthy adults and elderly individuals as well as the number of trials required to produce repeatable results. Methods Twenty-four healthy adults (21–63 years old) and 16 elderly adults (66–87 years) performed a voluntary rapid step execution following a tap on their heel while standing on a force platform under single and dual task conditions on three separate occasions. The first two tests were performed 30–60 minutes apart and the third test was performed a week later. Variables analyzed from the ground reaction force data included onset latency of step initiation (initiation phase), preparation and swing phases, foot-off and foot-contact times. Results Intraclass correlation coefficients (ICC(2,1)) were good to excellent across all parameters and test conditions for the pooled population and for elderly (0.74–0.92 and 0.62–0.88, respectively) except for the swing phase duration where lower values were seen (0.54–0.60 and 0.32–0.64 respectively). Values were similar under single and dual task conditions. Conclusion A voluntary step execution test, performed under single and dual task conditions especially foot-off and foot-contact times, is a reliable outcome measure that may be a useful tool to asses dynamic balance function for diagnostic purposes as well as clinical intervention trials.