Study 1 quantified the validity and repeatability of an automated on-line (ON) gas analysis system during sub-maximal loaded marching (LM) against that of the Douglas Bag (DB) approach. The 95% ratio Limits of Agreement (LoA) revealed the ON system systematically overestimated V02 by -16% (1.16 (x/-i-1.19). The Bland and Altman plots revealed DB repeatability was almost two-fold better than ON (-9% vs. -15%), thus the DB approach should be used subsequently to measure human expired gases. Study 2 investigated the difference between an LM maximal oxygen uptake protocol (LMp) versus a standard running protocol (Rp). The LMp V02max was lower than Rp (48.6 ± 4.3 ml·kg-I·min-I vs. 51.3 ± 4.0 ml·kg-I·min-I, P=0.001). Thus, the quantification of sub-maximal LM exercise intensity will be underestimated by -5% if derived from a running tiOzmax protocol. Study 3 investigated the repeatability of accepted and potential determinants of Loaded Marching Performance (LMP). The LoA revealed the repeatability of Loaded Marching Economy (LME) (0.98 (x/-i-1.09», V02max (1.01 (xl-i-1.07», upper body dynamic strength (1.01 (x/-i- 1.11», and anthropometric measures (1.00 (x/-i- 1.02» to (1.00 (x/-i- 1.07» was reasonable, but dynamic leg strength (1.06 (x/-i-1.14» and isometric strength (1.00 (x/-i-1.12» to (0.99 (x/-i-1.l6» were large. Study 4 established the determinants of 2.4 km LMP from a test battery performed at the beginning of British Army infantry training. The best mathematical model of LMP included the independent variables of LME (r=0.65), 2.4 km run time (r=0.42), and peak static lift strength (r=0.48). This explained 65% of the variation in LMP, and had a prediction error of ± 51 s. Mathematically, LME and 2.4 km run time exerted the greatest influence on LMP, whereas the influence of static lift strength on LMP was small. Study 5 investigated the physical and physiological responses of the established determinants of LMP during 24 weeks of British Army infantry training. Loaded marching performance improved 7.0% (900 s to 837 s, P=0.001), LME 9.6% (2.28 I·min- I to 2.06 I·min- I , P ...