There is still controversy about the reliability and prognostic value of QT interval dispersion because of interobserver and intraobserver variability. The authors aimed to study the effect of respiratory phases on QT dispersion. Sixty healthy volunteers (38 men, 22 women, mean age 25+/-3 years) from the medical staff comprised the study group. Electrocardiograms were recorded by the same technician with a rate of 50 mm/s during normal breathing, maximum inspiration and expiration. QT dispersion was defined as the difference between the maximal and minimal QT interval measurement occurring among any of the 12 leads. Corrected QT (QTc) interval was calculated according to Bazzet's formula. There were no significant differences between QTc max interval during maximum inspiration and expiration compared with those in normal breathing (409+/-20 ms vs 417+/-26 ms, p>0.05 and 412+/-18 vs 417+/-26 ms, p>0.05 respectively). QTc dispersion during maximum inspiration and expiration was significantly lower than that of normal breathing (36+/-8 ms vs 44+/-9 ms, p<0.003 and 32+/-7 vs 44+/-9 ms, p< 0.003, respectively). And the QTc dispersion during maximum expiration was also lower than that during maximum inspiration (p<0.01). QT dispersion magnitude is affected by the respiratory phases in healthy subjects and decreases during both maximum inspiration and expiration as compared with normal respiration.