Blood lactate responses are commonly employed for evaluation and prescription of training programmes. The purpose of the present studies was to examine the effects of dietary manipulations on both swim performance and on the relationship between blood lactate and swimming velocity. The first study engaged 8 subjects in a regimen to reduce muscle glycogen by a combination of diet and training. Subjects were monitored under a normal mixed diet comprising 53.6+/-7.8% carbohydrate (CHO) and under a decreased CHO condition (39.4+/-10.7% CHO) over 3 days. Mean swim performance decreased significantly (P<0.05) over 400 yards as a result of the carbohydrate reduction regimen. Mean blood lactates were reduced as a consequence of the glycogen depletion regimen following swims at 85% and 100% of maximum velocities. The swimming velocity corresponding to 4 mM blood lactate (V-4 mM) was altered from 0.67+/-0.04 m x s(-1) on a mixed diet to 0.70+/-0.05 m x s(-1) on a CHO-reduced diet. The second study employed 7 subjects in a regimen to enhance muscle glycogen stores. Subjects were monitored over 3 days under a normal diet (52.7+/-4.4% CHO) and on a separate occasion under an increased CHO intake (59.2+/-3.7% CHO). In contrast to the first study, mean swim performance improved over 100 yards and 400 yards (P<0.05). Mean blood lactates were evaluated after the carbohydrate-rich regimen at both 85% and 100% swim velocities (P<0.05). The mean swim velocity associated with V-4mM was paradoxically reduced from 0.69+/-0.05 to 0.67+/-0.04 m x s(-1) as a result of the increased CHO condition. The results indicate that a moderate reduction in CHO intake alters swimming performance adversely whereas a moderate elevation in CHO intake above the normal diet improves performance. The dietary manipulations affected the response of blood lactate to both submaximal and maximal swimming velocities. The observations highlight the limitations of applying lactate response curves to swim training.