Interval exercise (IE) interventions drive superior physiological adaptations compared to continuous exercise (Wisløff et al., 2007). However, identification of which IE protocol most effectively promotes physiological adaptations is unclear as the IE work rate (WR) and work:recovery durations interact with oxygen uptake (VO2) dynamics to influence the intensity domain in which an individual is working (Turner et al., 2006). Therefore, we assessed whether an individualised computational model along with measures from two standard exercise tests would allow prediction of the WRs that would place subjects in a desired exercise intensity domain for a variety of IE protocols. Eight healthy males (mean±SD age 21±1 y, height 178±6 cm, mass 75±8 kg) performed a ramp incremental test to the limit of tolerance on a cycle ergometer. Lactate threshold (LT), an index of critical power (CP) and peak VO2 (VO2peak) were estimated from these data to obtain thresholds for the exercise intensity domains. On a separate visit, subjects performed a moderate-intensity step protocol to measure baseline VO2 (VO2base), gain (ΔVO2/ΔW) and phase 2 VO2 kinetics (τVO2). These data were used to individually parameterise a computational model of gas exchange and circulatory dynamics (Benson et al., 2013). The model was used to predict the WRs that would place each subject in the moderate, heavy and very heavy exercise intensity domains for work:recovery durations of 15:15, 30:30, 60:60 and 120:120 s (i.e. 12 WRs per subject), with these each performed for up to 22 min over 12 subsequent visits. Measured IE breath-by-breath VO2 was then compared to model-predicted VO2 outputs. Measured parameters were: VO2base, 0.7±0.07 (range: 0.61-0.8) l.min-1; τVO2, 29.9±6.0 (20.7-39.3) s; LT, 1.82±0.22 (1.58-2.28) l.min-1; CP index, 2.77±0.40 (2.35-4.05) l.min-1, and VO2peak, 3.58±0.42 (3.09-4.53) l.min-1, highlighting the large inter-variability in model inputs and intensity domain thresholds (LT range 39-59 % VO2peak; CP index range 56-90 % VO2peak). Mean error between measured and model-predicted VO2 responses ranged from 0.11±0.02 l.min-1 for the moderate intensity 15:15 s protocol to 0.32±0.11 l.min-1 for the very heavy intensity 15:15 s protocol. However, peak metabolic disturbance during IE (i.e. the peak of the VO2 oscillation in each work bout) did not exceed LT (for moderate-intensity IE protocols) and the CP index (for heavy intensity IE protocols), and in 59% of very heavy intensity IE protocols the limit of tolerance was reached. Thus, a computational model and data from two standard exercise tests can predict individualised WRs that place subjects in the desired exercise intensity domain for a range of IE protocols. This therefore provides a methodology to control for exercise intensity when investigating how IE can be optimised to maximise physiological adaptations.