The correlations found in this investigation are in agreement with the results of Sirtori et al. (1993) and of two other similar studies (Jacobs et al., 1983; Duggan and Tebbutt, 1990). Jacobs et al. (1983) found a significant relationship (r =−0.95) between [La−] at the end of a single submaximal cycling workload at 200 W and OBLA. A similar correlation (r=−0.95) was reported by Duggan and Tebbott (1990) between OBLA vclocity and [La−] measured during submaximal running at 12 km·h−1. However, in our study correlations were lower (from r=0.62 to r= −0.76) than those reported in these studies. The reason is not clear but could be related to different laboratory protocols, environmental variables during field tests or simply due to peculiar characteristics of our group of soccer players. Despite being significant, the correlation coefficients found do not seem to support the use of OBLA velocity estimation as suggested by Sirtori et al. (1993). In fact, the moderate correlations of this study can be sufficient to validate the use of [La−]FIELD in evaluating aerobic changes but they are not sufficiently high to allow the prediction of laboratory lactate thresholds with acceptable accuracy (SEE from 7.1% to 10.2%). Even using the regression equation proposed by Sirtori et al. (1993), correlations between lactate thresholds determined under laboratory conditions and estimated OBLA velocity from [La−]FIELD were moderate (from r=−0.65 tor=−0.78).