Sprint interval training (SIT) has emerged as a time efficient alternative to conventional endurance exercise, inducing beneficial physiological adaptations. Nutritional support, including exercise in the fasted state, to optimise the response to SIT remains poorly characterised. The purpose of this study was to compare the effects of fasted (FAST) and CHO-fed (FED) SIT on exercise performance in recreationally active males and to describe the inter-individual variation in response to SIT. This study received local ethical approval and was conducted in line with the principles of the World Medical Association’s Declaration of Helsinki. Healthy, untrained, recreationally active (VO2 max < 50 ml.kg.min-1) males (n=17) participated in a 3 week SIT intervention, comprising 9 exercise sessions performed in either a fasted or CHO-fed condition. Participants were randomised to FAST and FED groups at baseline. For each exercise session, participants arrived to the lab following an overnight fast (≥10 h) and ingested either 0.33 g.kg-1 body mass (BM) of non-caloric placebo (FAST), or 0.91 g.kg-1 BM maltodextrin (FED). SIT consisted of 4-6 X 30 s “all out” cycle sprints at a resistance of 7.5 % BM, interspersed with 4 min active recovery. Participants completed a 30 s Wingate test and a 20 min cycle performance test at baseline and 72-96 h following the final SIT session. (Non-)Responders were identified using twice the typical error (TE) of measurement for Wingate peak power (PP; 2 X TE = 1.58 W.kg-1), minimum power (MP; 2 X TE = 0.81 W.kg-1) and mean power (AP; 2 X TE = 0.39 W.kg-1), and for mean power during the cycle performance test (CP; 2 X TE = 10.07 W). Independent t-tests determined between-group differences in change in exercise performance with data reported as mean ± SD. There was no significant difference between FAST and FED groups ΔPP (0.6 ± 1.2 vs. 1.0 ± 1.0 W.kg-1), ΔMP (0.5 ± 0.6 vs. 0.1 ± 0.8 W.kg-1), ΔAP (0.3 ± 0.5 vs. 0.3 ± 0.4 W.kg-1) or ΔCP (22.6 ± 11.3 vs. 19.3 ± 20.3 W; all p > 0.05). Non-responders for PP (14/18, 77 %), MP (14/18, 77 %), AP (11/18, 61 %), and CP (7/18, 38 %) were identified across both groups. Global non-responders who did not improve any performance measure were also identified (3/18, 17 %). A greater number of responders to CP were identified in the FAST compared to FED (FAST: 88%; FED 44%). The number of responders between groups was similar for PP (FAST: 25%; FED 22%), AP (FAST: 50%; FED 33%) and MP (FAST: 25%; FED 22%). In conclusion, these findings indicate no significant difference in performance adaptations between FAST and FED SIT, however, heterogeneity is evident in the individual adaptations to SIT in measures of anaerobic and aerobic performance in both FAST and FED.