While the physiological adaptations following endurance training are relatively well understood, in swimming there is a dearth of knowledge regarding the metabolic responses to interval training (IT). The hypothesis tested predicted that two different endurance swimming IT sets would induce differences in the total time the subjects swam at a high percentage of maximal oxygen consumption (VO(2)max). Ten trained triathletes underwent an incremental test to exhaustion in swimming so that the swimming velocity associated with VO(2)max (vVO(2)max) could be determined. This was followed by a maximal 400-m test and two intermittent sets at vVO(2)max: (a) 16 x 50 m with 15-s rest (IT(50)); (b) 8 x 100 m with 30-s rest (IT(100)). The times sustained above 95% VO(2)max (68.50 +/- 62.69 vs. 145.01 +/- 165.91 sec) and 95% HRmax (146.67 +/- 131.99 vs. 169.78 +/- 203.45 sec, p = 0.54) did not differ between IT(50) and IT(100)(values are mean +/- SD). In conclusion, swimming IT sets of equal time duration at vVO(2)max but of differing work-interval durations led to slightly different VO(2)and HR responses. The time spent above 95% of VO(2)max was twice as long in IT(100) as in IT (50), and a large variability between mean VO(2)and HR values was also observed.