We investigated the validity of different lactate and ventilatory threshold methods, to estimate heart rate and power output corresponding with the maximal lactate steady-state (MLSS) in elite cyclists. Elite cyclists (n = 21; 21 +/- 0.4 y; VO2peak, 5.4 +/- 0.2 l x min (-1)) performed either one (n = 10) or two (n = 11) maximal graded exercise tests, as well as two to three 30-min constant-load tests to determine MLSS, on their personal race bicycle which was mounted on an ergometer. Initial workload for the graded tests was 100 Watt and was increased by either 5 % of body mass (in Watt) with every 30 s (T30 s), or 60 % of body mass (in Watt) with every 6 min (T6min). MLSS was defined as the highest constant workload during which lactate increased no more than 1 mmol x l (-1) from min 10 to 30. In T30 s and T6 min the 4 mmol (TH-La4), the Conconi (TH-Con) and dmax (TH-Dm) lactate threshold were determined. The dmax lactate threshold was defined as the point that yields the maximal distance from the lactate curve to the line formed by the lowest and highest lactate values of the curve. In T30 s also ventilatory (TH-Ve) and Vslope (TH-Vs) thresholds were calculated. Time to exhaustion was 36 +/- 1 min for T30 s versus 39 +/- 1 min for T6 min. None of the threshold measures in T30 s, except TH-Vs (r2 = 0.77 for heart rate) correlated with either MLSS heart rate or power output. During T6 min, power output at TH-Dm was closely correlated with MLSS power (r2=0.72). Low correlations were found between MLSS heart rate and heart rate measured at TH-Dm (r2=0.46) and TH-La4 (r2=0.25), respectively, during T6 min. It is concluded that it is not possible to precisely predict heart rate or power output corresponding with MLSS in elite cyclists, from a single graded exercise test causing exhaustion within 35-40 min. The validity of MLSS predicted from an incremental test must be verified by a 30-min constant-load test.