This study investigated which exercise mode (continuous or sprint interval) is more effective for improving insulin sensitivity. Ten young, healthy men underwent a non-exercise trial (CON) and 3 exercise trials in a cross-over, randomized design that included 1 sprint interval exercise trial (SIE; 4 all-out 30-s sprints) and 2 continuous exercise trials at 46% VO2peak (CELOW) and 77% VO2peak (CEHIGH). Insulin sensitivity was assessed using intravenous glucose tolerance test (IVGTT) 30 min, 24 h and 48 h post-exercise. Energy expenditure was measured during exercise. Glycogen in vastus lateralis was measured once in a resting condition (CON) and immediately post-exercise in all trials. Plasma lipids were measured before each IVGTT. Only after CEHIGH did muscle glycogen concentration fall below CON (P<0.01). All exercise treatments improved insulin sensitivity compared with CON, and this effect persisted for 48-h. However, 30-min post-exercise, insulin sensitivity was higher in SIE than in CELOW and CEHIGH (11.5±4.6, 8.6±5.4, and 8.1±2.9 respectively; P<0.05). Insulin sensitivity did not correlate with energy expenditure, glycogen content, or plasma fatty acids concentration (P>0.05). After a single exercise bout, SIE acutely improves insulin sensitivity above continuous exercise. The higher post-exercise hyperinsulinemia and the inhibition of lipolysis could be behind the marked insulin sensitivity improvement after SIE.