The purpose of this study was to examine the physiological effects of 3000-m swimming on subsequent 3-h cycling time trial performance in ultraendurance triathletes. Eight highly trained ultraendurance triathletes [mean (SEM) age 34 (2) years, body fat 12.5 (0.8)%, maximum oxygen consumption 63.2 (2.1) ml x kg(-1) x min(-1)] completed two randomly assigned trials 1 week apart. The swim/bike trial (SB) involved 3000 m of swimming [min:s 52:28 (1:48)] immediately followed by a 3-h cycling performance at a self-selected time-trial pace. The control trial (CON) consisted of an identical 3-h cycling time trial but without prior swimming. Subjects consumed an 8% carbohydrate (CHO)/electrolyte beverage during both trials at the rate of 60 g CHO x h(-1) and 1 l x h(-1). No significant differences were evident between CON and SB on the dependent measures (CON vs. SB): power output [W, 222 (14) W vs. 212 (13) W], heart rate [fc, 147 (5) beats x min(-1) vs. 143 (4) beats min(-1); %fcmax 80.0 (1.6)% vs. 78.4 (1.5)%], oxygen uptake [3.10 (0.12) l x min(-1) vs. 2.97 (0.15) l x min(-1)], minute ventilation [82.5 (4.4) l x min(-1) vs. 77.3 (3.7) l x min(-1)], rating of perceived exertion [14.6 (0.4) vs. 14.0 (0.1)], blood lactate [6.1 (0.5) mmol x l(-1) vs. 4.8 (0.5) mmol x l(-1)], and blood glucose [5.0 (0.2) mmol x l(-1) vs. 5.3 (0.1) mmol x l(-1); all non-significant at the P>0.05 level]. However, the CON respiratory exchange ratio was significantly greater than for SB [0.91 (0.01) vs. 0.89 (0.01); P<0.05], suggesting that the SB trial required a greater reliance on lipid as a fuel substrate. Hence, the main finding in the present study was that 3000 m of swimming had no significant performance effect (in terms of W) on subsequent 3-h cycling performance in ultraendurance triathletes.