While lactate threshold (LT) is a common laboratory test and is often used to design training plans and monitor fitness changes, there is no consensus as to which determination method should be used. Three of the most common are the break point (BREAK), 1 mmol.L-1 over baseline (+1 mmol), and reaching 4 mmol.L-1 (onset of blood lactate accumulation, OBLA). Currently, it is not clear if different determination methods could yield varying results in different populations. PURPOSE: We compared 3 most commonly used LT methods in trained cyclists and in athletes (soccer players) who were not accustomed to cycling exercise. We hypothesized that the BREAK and +1 mmol methods would yield similar results in both populations. METHODS: LT data were analyzed from 41 athletes [18 trained cyclists (15 males [m], 3 females [f]) and 23 female soccer players]. Tests were performed on a cycle ergometer using 5 min stages starting at 70 W (m)/50 W (f). Work rates were increased by 25 W (m)/15 W (f) for the first 3-4 stages, and by 15 W (m)/10 W (f) for the last 2-3 stages. Blood samples were obtained in the last min of each stage, and blood lactate was analyzed using a Lactate Plus device. For determinations of LT, 3 trained investigators independently analyzed the plots. RESULTS: In cyclists, LTs using BREAK and +1 mmol (247±48 W vs. 250±50 W, p=0.52) were not different but were significantly lower than that obtained with OBLA (270±54 W). Correlational analyses indicate that LT using BREAK and +1 mmol were strongly related (R=0.99). Associations were strong between +1 mmol and OBLA (R=0.96) and between BREAK and OBLA (R=0.96). In non-cyclists, LTs obtained with all 3 methods were significantly different (BREAK: 125±13 W; +1 mmol: 130±11 W; OBLA, 134±13 W; all p<0.04), although BREAK and +1 mmol were strongly associated (R=0.98). The associations between different LT methods were much weaker in non-cyclists (1 mmol and OBLA: R=0.90; BREAK and OBLA: R=0.85) compared with cyclists. CONCLUSIONS: Break point and +1 mmol.L-1 methods yield comparable results in trained cyclists but not in non-cyclists. Caution should be used when interpreting LT results obtained from different determination methods.