Stepping thresholds, i.e. the maximum perturbation one can withstand without taking a step, predict falls in older people. This ability requires fast central processing of sensory information followed by rapid execution of adequate motor responses, both of which are affected by age. However, there is limited evidence on their combined effect on stepping thresholds.Are cognitive and motor speeds important for stepping thresholds and do they interact, allowing for compensation?Two-hundred forty-two people (mean age: 80 years, standard deviation 4; 110 women) underwent a series of waist-pulls of increasing magnitude to assess stepping thresholds in anterior, posterior and mediolateral directions. Cognitive function was assessed as simple hand reaction time and trail making test performance, and muscle function was assessed as isometric peak and rate of torque development of dominant leg muscles. Principal component analysis reduced these variables to four factors: peak muscle strength, muscle torque development speed (motor speed), executive function and central processing speed (cognitive speed). These factors were used in univariable and multivariable regression models to determine their association with stepping thresholds.Faster central processing speed (beta:2.69; 95 %CI:1.49-3.88) and faster muscle torque development speed (beta:2.60, 95 %CI:0.63-4.57) were associated with higher stepping thresholds. These associations remained in a multivariable model. No interaction was found between cognitive and motor speed on stepping thresholds (p = 0.602).Central processing speed and muscle torque development speed affect stepping thresholds independently from each other and may both be important age-related motor impairment targets for preventing falls in older people.