Abstract

Repeated practice produces both motor automaticity (action fluency) and inflexible habitual responding (non-deliberate actions). Although traditionally assumed to reflect a unified automaticity process, their relationship remains largely untested. Using a novel paradigm designed to enhance and measure habit expression (assessed via outcome devaluation) alongside motor automaticity (indexed by established markers including timing consistency and execution speed), we found a counterintuitive inverse relationship: individuals with greater motor automaticity showed reduced habitual responding. This finding was replicated in a pre-registered confirmatory study and generalized across three independent datasets spanning different paradigms, training durations, action types, reinforcement schedules, and laboratories (total N=1,065 across all five datasets). The inverse relationship converged across multiple motor automaticity indices. Additionally, we provide a window into habit consolidation, showing that motor automaticity during habitual errors is preserved after extensive training but disrupted after limited training. Our findings dissociate execution automaticity from selection automaticity (how actions are performed versus whether and which actions are initiated), challenging prevailing assumptions in neurocognitive accounts of learning. They revise our understanding of how practice shapes behavior, with implications or distinguishing adaptive skill from maladaptive rigidity in healthy and clinical populations.