The purpose of this study was to determine whether twitch potentiation (a transient augmentation of twitch tension following vigorous muscular contraction) occurs in the human diaphragm in vivo. Six healthy subjects were studied. To determine whether twitch potentiation occurs, the subjects attempted to maximally activate their diaphragm by performing the combined Mueller-expulsive maneuver with visual feedback (Pdi max maneuver). Twitches were obtained before, at 10 s after the transdiaphragmatic pressure (Pdi) maneuver, and at intervals over the ensuing 10 min. We also determined whether twitch potentiation would occur following submaximal voluntary diaphragmatic contractions (33 and 66% of Pdi max). In addition, we examined whether repeated voluntary contractions would result in greater twitch potentiation compared with that observed after a single voluntary contraction. Twitch potentiation was observed in every subject. The number of maximal voluntary contractions (MVC) (one, two, or four) had no significant effect on the degree of twitch potentiation. The increase in twitch amplitude (expressed as a percentage of the control value) averaged 63 +/- 35% (SD)(pooled data from one, two, and four contraction trials). Twitch potentiation decayed in a monoexponential fashion (r = 0.99) with a time constant of 125 s (95% Cl = 100 to 160 s). Twitch potentiation was also observed after submaximal voluntary diaphragmatic contractions. Again, the number of voluntary contractions (one or four) had no significant effect on the degree of twitch potentiation. After submaximal diaphragmatic contractions of 66% of Pdi max, the degree of twitch potentiation was not significantly different, 61 +/- 36% (pooled data from one and four contraction trials) from that observed following the maximal voluntary contraction maneuvers.(ABSTRACT TRUNCATED AT 250 WORDS)