Triathlon running is affected by prior cycling and power output during triathlon cycling is variable in nature. We compared constant and triathlon-specific variable power cycling and their effect on subsequent submaximal running physiology. Nine well-trained male triathletes (age 24.6 ± 4.6 years, VO_ 2peak 4.5 ± 0.4 L · min-1 ; mean ± SD) performed a submaximal incremental run test, under three conditions: no prior exercise and after a 1 h cycling trial at 65% of maximal aerobic power with either a constant or a variable power profile. The variable power protocol involved multiple 10–90 s intermittent efforts at 40–140% maximal aerobic power. During cycling, pulmonary ventilation (22%, ±14%; mean; ±90% confidence limits), blood lactate (179%, ±48%) and rating of perceived exertion (7.3%, ±10.2%) were all substantially higher during variable than during constant power cycling. At the start of the run, blood lactate was 64%, ±61% higher after variable compared to constant power cycling, which decreased running velocity at 4 mM lactate threshold by 0.6, ±0.9 km · h-1 . Physiological responses to incremental running are negatively affected by prior cycling and, to a greater extent, by variable compared to even-paced cycling. Testing and training of triathletes should account foe higher physiological cost of triathlon-specific cycling and its effect on subsequent running.