In cool conditions, physiologic markers accurately predict endurance performance, but it is unclear whether thermal strain and perceived thermal strain modify the strength of these relationships. This study examined the relationships between traditional determinants of endurance performance and time to complete a 5-km time trial in the heat. Seventeen club runners completed graded exercise tests (GXT) in hot (GXTHOT; 32°C, 60% RH, 27.2°C WBGT) and cool conditions (GXTCOOL; 13°C, 50% RH, 9.3°C WBGT) to determine maximal oxygen uptake (V̇O2max), running economy (RE), velocity at V̇O2max (vV̇O2max), and running speeds corresponding to the lactate threshold (LT, 2 mmol.l(-1)) and lactate turnpoint (LTP, 4 mmol.l(-1)). Simultaneous multiple linear regression was used to predict 5 km time, using these determinants, indicating neither GXTHOT (R(2) = 0.72) nor GXTCOOL (R(2) = 0.86) predicted performance in the heat as strongly has previously been reported in cool conditions. vV̇O2max was the strongest individual predictor of performance, both when assessed in GXTHOT (r = -0.83) and GXTCOOL (r = -0.90). The GXTs revealed the following correlations for individual predictors in GXTHOT; V̇O2maxr = -0.7, RE r = 0.36, LT r = -0.77, LTP r = -0.78 and in GXTCOOL; V̇O2maxr = -0.67, RE r = 0.62, LT r = -0.79, LTP r = -0.8. These data indicate (i) GXTHOT does not predict 5 km running performance in the heat as strongly as a GXTCOOL, (ii) as in cool conditions, vV̇O2max may best predict running performance in the heat.