The purpose of this study was to determine the relationship between actual time to exhaustion or time limit (ATLIM) during bicycle ergometry and predicted time to exhaustion (PTLIM) from the Critical Power (CP) test. fourteen males (means +/- SD = 22.36 +/- 2.13 years) volunteered as subjects for this investigation. The subjects visited the laboratory on seven occasions separated by at least 24 h. The first two visits were used for the determination of CP; during the remaining sessions the subjects rode a Monarch bicycle ergometer at power loadings of CP - 20%, CP, CP + 20%, CP + 40% and CP + 60% for the determination of ATLIM. Theoretically, power loadings less than or equal to CP can be maintained indefinitely without exhaustion and the PTLIM for power loadings greater than CP can be estimated from the results of the CP test. The accuracy of the CP test for estimating the time to exhaustion during bicycle ergometry was determined by comparing ATLIM to PTLIM using correlation coefficients, standard error of estimates and related t-tests. The results of this study indicated that there were no significant (p greater than 0.05) differences between ATLIM and PTLIM for power loadings greater than CP (ATLIM vs PTLIM at CP + 20% = 8.19 +/- 3.90 vs 7.13 +/- 2.69 min, t = 2.106, r = 0.893, SEE = 1.21 min; CP + 40% = 3.60 +/- 1.37 vs 3.46 +/- 1.18 min, t = 0.842, r = 0.882, SEE = 0.556 min; CP + 60% = 2.36 +/- 0.95 vs 2.32 +/- 0.79 min; t = 0.328 r = 0.841, SEE = 0.428 min).(ABSTRACT TRUNCATED AT 250 WORDS)