Controversy persists regarding the mechanism underlying the lactate threshold. It has recently been argued that there is in fact no "threshold" and that blood lactate increases as a continuous function during exercise (Hughson J. Appl. Physiol. 62:1975-1981, 1987). In comparing continuous and threshold models, questions have been raised regarding the ramp rate, data sampling, and the mathematical models employed (Morton J. Appl. Physiol. 67:885-888, 1989). To address some of these concerns, we evaluated 61 subjects (mean age 45 +/- 15), who underwent maximal ramp treadmill tests with the ramp rate individualized such that test duration was approximately 10 min for each subject. The relationship between changes in blood lactate and oxygen uptake were evaluated using a modification of the log-log transformation model described by Beaver (J. Appl. Physiol. 59:1936-1940, 1985) and a continuous exponential plus constant model described by Hughson et al. (J. Appl. Physiol. 62:1975-1981, 1987). Model fitting, using mean squared error (MSE) and coefficient of determination (CD) for each method were as follows: [table: see text] The modified log-log model had a better fit as indicated by the lower MSE and higher CD, suggesting the change in lactate was better described by this model. However, the differences were so slight as to suggest: 1) a meaningful difference does not exist between the two; or 2) these methods may not be capable of detecting a difference, if one exists.