The article presents a stochastic optimized-submovement model to explain Fitts' law, which describes the logarithmic trade-off between movement speed and accuracy in rapid aimed movements. According to the model, an aimed movement involves a primary submovement and an optional secondary corrective submovement. These submovements are programmed to minimize total movement time while maintaining high accuracy. The model suggests that submovements are optimized by adjusting the average magnitudes and durations of noisy neuromotor force pulses. The model explains various findings in motor performance literature and is supported by two experiments on rapid wrist rotations. The results show that primary submovement durations and endpoint variability align with the model's predictions. The model also accounts for the effects of visual feedback on movement accuracy and time. The deterministic iterative-corrections model, which assumes a series of discrete submovements guided by feedback, has limitations in explaining submovement endpoint variability and error rates. The stochastic optimized-submovement model, in contrast, incorporates neuromotor noise and optimizes submovement durations to minimize total movement time. The model predicts a square-root function for the speed-accuracy trade-off, which closely approximates Fitts' law. The model also explains the relative frequency of secondary submovements and the effects of neuromotor noise on submovement endpoints. The model provides a framework for understanding motor performance and links it to research on sensation, perception, and cognition. The model's predictions are supported by empirical data and offer a more comprehensive explanation of motor performance than previous models.The article presents a stochastic optimized-submovement model to explain Fitts' law, which describes the logarithmic trade-off between movement speed and accuracy in rapid aimed movements. According to the model, an aimed movement involves a primary submovement and an optional secondary corrective submovement. These submovements are programmed to minimize total movement time while maintaining high accuracy. The model suggests that submovements are optimized by adjusting the average magnitudes and durations of noisy neuromotor force pulses. The model explains various findings in motor performance literature and is supported by two experiments on rapid wrist rotations. The results show that primary submovement durations and endpoint variability align with the model's predictions. The model also accounts for the effects of visual feedback on movement accuracy and time. The deterministic iterative-corrections model, which assumes a series of discrete submovements guided by feedback, has limitations in explaining submovement endpoint variability and error rates. The stochastic optimized-submovement model, in contrast, incorporates neuromotor noise and optimizes submovement durations to minimize total movement time. The model predicts a square-root function for the speed-accuracy trade-off, which closely approximates Fitts' law. The model also explains the relative frequency of secondary submovements and the effects of neuromotor noise on submovement endpoints. The model provides a framework for understanding motor performance and links it to research on sensation, perception, and cognition. The model's predictions are supported by empirical data and offer a more comprehensive explanation of motor performance than previous models.