Joint biomechanics and spatiotemporal gait parameters change with age or disease and are used in treatment decision-making. Research question: To investigate whether kinematic predictors of spatiotemporal parameters during gait differ by age in healthy individuals. Methods: We used an open dataset with the gait data of 114 young adults (M = 28.0 years, SD = 7.5) and 128 older adults (M = 67.5 years, SD = 3.8) walking at a comfortable self-selected speed. Linear regression models were developed to predict spatiotemporal parameters separately for each group using joint kinematics as independent variables. Results: In young adults, knee flexion loading response and hip flexion/extension were the common predictors of gait speed; hip flexion and hip extension contributed to explaining the stride length; hip flexion contributed to explaining the cadence and stride time. In older adults, ankle plantarflexion, knee flexion loading response, and pelvic rotation were the common predictors of the gait speed; ankle plantarflexion and knee flexion loading response contributed to explaining the stride length; ankle plantarflexion loading response and ankle plantarflexion contributed to explain the cadence, stride width and stride time. Significance: Our results suggest that the ability of joint kinematic variables to estimate spatiotemporal parameters during gait differs by age in healthy individuals. Particularly in older adults, ankle plantarflexion was the common predictor of the spatiotemporal parameters, suggesting the importance of the ankle for gait parameters in this age group. This provides insight for clinicians into the most effective evaluation and has been used by physical professionals in prescribing the most appropriate exercises to attenuate the effects produced by age-related neuromuscular changes.