Laboratory studies with competitive athletes often use graded exercise protocols to elicit physiologic responses. This pattern of power output is different than ordinarily employed by athletes during competition. To understand the physiologic responses during competition, we studied 24 athletes (speed skaters, cyclists, triathletes) during simulated competition, a 5-km time trial on a racing bicycle attached to a windload simulator, and during cycle ergometer graded exercise testing (N = 8). During the time trial the velocity pattern was similar to real world competitions, and the subjects indicated that the time trial was perceptually similar to competition. Physiologic responses were of significantly greater magnitude vs graded exercise (VO2max: 3.46 +/- 0.73 vs 3.27 +/- 0.79 l.min-1; VEmax: 138 +/- 27 vs 119 +/- 22 l.min-1; HRmax 184 +/- 11 vs 175 +/- 11 beats x min-1; HLa 14.8 +/- 3.7 vs 11.9 +/- 2.1 mM). All physiologic measures increased steadily throughout the time trial (km 0, 1, 2, 3, 4, and 5: VO2 = 1.03, 2.95, 3.42, 3.69, 3.82, and 3.92 l.min-1; HR = 93, 175, 181, 185, 189, and 194 beats.min-1; VE = 31, 99, 120, 129, 145, and 156 l.min-1; HLa = 2.9, 5.6, 7.2, 9.2, 10.6, and 13.5 mM). In six subjects (speed skaters), the peak values observed during time trial for HR (188 +/- 6 vs 191 +/- 5 beats.min-1) and HLa (16.4 +/- 3.1 vs 17.0 +/- 4.2 mM) were not significantly different than observed during real world competition.(ABSTRACT TRUNCATED AT 250 WORDS)