The specificity of three maximal cycling sprint tests as a measure of anaerobic lactacid capacity was determined in nine highly trained male cyclists when they performed 10-, 30-, 40-, and 60-s tests on a modified Repco wind-braked cycle ergometer. Peak power (PP), percent power loss (% PO), total work done (TW), and peak blood lactate (PHLa) were determined for each test. The cyclists also performed a 1000-m time trial under competition conditions during which 200-m split times, total time (TT), and peak post-competition blood lactate (TTPHLa) were recorded. While there was no statistically significant difference between the peak blood lactate of the 30-, 40-, and 60-s tests, peak blood lactate achieved after the 1000-m time trial was significantly greater than those after the cycle ergometer tests. Although there were high intercorrelations (0.88-0.99) between the anaerobic power and capacity indices of the laboratory tests, only the PP and TW achieved during the 60-s test correlated significantly (P less than 0.05) with TT. The data suggest that when assessing the anaerobic power and capacity of elite 1000-m time trial cyclists, a cycle ergometer test duration of at least 60 s should be employed.