Self‐talk is a psychological skill that benefits motor performance by controlling and organizing performers’ thoughts. While the behavioral effects of self‐talk are clear, research on the mechanisms underpinning the effects of different modes of self‐talk is sparse. To address this issue, we propose and test a psychophysiological model of the effects of self‐talk on motor performance. Forty golf novices practiced a golf putting task while using either instructional or motivational self‐talk preceding each putt. We measured performance (radial error), technique (club kinematics and muscle activity), cardiac activity (heart‐rate and event‐related heart‐rate change), as well as electroencephalographic alpha power and connectivity in a randomized (group: instructional self‐talk, motivational self‐talk) experimental design. Instructional self‐talk promoted superior technique and was associated with greater parietal alpha power and weaker connectivity between frontal and parietal electrodes and all other scalp sites, possibly indicative of increased top‐down control of action. These findings provide initial evidence for an information‐processing mechanism underlying the benefits of instructional self‐talk. They also cast doubt on the validity of left‐frontotemporal connectivity as a measure of verbal‐analytic processing during motor tasks. Motivational self‐talk led to increased heart‐rate and reduced event‐related heart rate variability, suggesting an effort‐based mechanism to explain the benefits of motivational self‐talk. Our study represents the most complete multi‐measure investigation of self‐talk to date. We hope that our psychophysiological model of self‐talk will encourage researchers to move beyond the exclusive reliance on behavioral and self‐report measures to discover the mechanisms underlying the benefits of self‐talk for performance.