Whilst group level changes in neural drive and muscle hypertrophy have been widely reported after strength training (Tillin et al., 2011, 2012; Erskine et al., 2012), the contributions of these adaptations to individual changes in strength are poorly understood. The purpose of this study was to assess the contribution of underlying physiological adaptations (neural, intrinsic contractile properties, muscle size and architecture) to the functional changes in explosive and maximal strength following training. Thirty-five healthy young males completed explosive strength training (EST, n=12), conventional strength training (CST, n=13), or control (CON, n=10) for 12 wks. Training involved 4 x 10 knee extension repetitions (x3/wk): contracting “as fast and hard as possible” for ~1 s (EST); or gradually increasing to 75% of maximal voluntary torque (MVT) before holding for 3 s (CST). Functional and physiological changes were measured pre and post. Knee extension torque (T) and quadriceps EMG were measured during maximum voluntary (MVT and EMG@MVT) as well as explosive voluntary (EMG0-50, 0-100, 0-150) and evoked contractions (T at 50 ms increments, T50, 100, and 150). Quadriceps muscle volume (QVOL, via MRI), pennation angle and fascicle length (via ultrasound recordings) were also determined. Pearson’s product moment bivariate correlations and, when multiple predictor variables were correlated with the outcome, stepwise multiple linear regressions were calculated between strength changes (ΔMVT, ΔT50, ΔT100, ΔT150) and the changes in physiological predictor variables. ΔMVT was correlated with ΔEMG@MVT (r=0.553, P=0.001) and ΔQVOL (r=0.608, P<0.001) but none of the other predictor variables (r≤0.216, P≥0.213). ΔT50 and ΔT100 were only correlated with ΔEMG0-50 (r=0.730, P<0.001) and ΔEMG0-100 (r=0.561, P<0.001), respectively. ΔT150 was correlated with ΔEMG0-150 (r=0.667, P<0.001) and ΔMVT (r=0.550, P=0.001). Stepwise regression with ΔQVOL (37%) and ΔEMG@MVT (17%) explained a combined 54% of the variance in the ΔMVT. ΔEMG0-150 (44%) and ΔMVT (10%) in combination explained 54% of the variation in ΔT150. In conclusion, improvements in MVT were explained primarily by ΔQVOL, with a smaller contribution from changes in neural drive. In contrast, changes in early phase explosive torque production (0-100 ms) were explained exclusively by changes in neural drive. Changes in late phase explosive force production were also largely explained by changes in neural drive but with a contribution from changes in MVT.