MORGAN, D. W., P. E. MARTIN, G. S. KRAHENBUHL, and F. D. BALDINI. Variability in running economy and mechanics among trained male runners. Med. Sci. Sports Exerc., Vol. 23, No. 3, pp. 378–383, 1991. Data from two studies were analyzed to quantify intraindividual variability and reliability in running economy (RE) and mechanics. Following 30–60 min of treadmill accommodation, stride-to-stride and day-to-day biomechanical stability were assessed in 31 male runners (studies 1 and 2) who performed two level treadmill runs (3.33 m.s-1) at the same time of day, in the same footwear, and in a nonfatigued state. Under the same testing conditions, daily stability in RE was determined in 17 of the 31 subjects (study 2). RE demonstrated high day-to-day reliability (r = 0.95), and the mean coefficient of variation in RE was 1.32% (range = 0.30–4.40%). Stride-to-stride reliability for temporal (T), kinematic (KNM), and kinetic (KIN) measures was very high (mean r = 0.96; range = 0.91–0.99), but day-to-day reliability was lower for KIN (mean r = 0.67; range = 0.28–0.88) compared with T and KNM (mean r = 0.91; range = 0.68–0.98). Further analyses revealed no significant (P ≤ 0.05) mean stride-to-stride differences for any biomechanical variable, and only three of 22 variables (peak resultant velocity of the ankle joint, step length, and swing time) demonstrated statistically significant day-to-day differences. Viewed in concert, these results suggest that, if the testing environment is controlled, multiple trials are not required to obtain stable measures of running economy and basic mechanical characteristics in trained male runners if group data from adequate sample sizes are considered. However, if individual records are scrutinized or if small sample sizes cannot be avoided, at least two measures of individual performances should be secured.