Quadriceps muscle and fibre cross-sectional areas (CSA), torque and neural activation were studied in seven healthy males during 6 months of weight training on alternate days with six series of eight unilateral leg extensions at 80% of one repetition maximum. After training, the quadriceps cross-sectional area increased by 18.8 +/- 7.2% (P < 0.001) and 19.3 +/- 6.7% (P < 0.001) in the distal and proximal regions respectively, and by 13.0 +/- 7.2% (P < 0.001) in the central region of the muscle. Hypertrophy was significantly different between and within the four constituents of the quadriceps. Biopsies of the vastus lateralis at mid-thigh did not show any increase in mean fibre cross-sectional area. Maximum isometric voluntary torque increased by 29.6 +/- 7.9%-21.1 +/- 8.6% (P < 0.01-0.05) between 100 degrees and 160 degrees of knee extension, but no change in the optimum angle (110 degrees-120 degrees) for torque generation was found. A 12.0 +/- 10.8% (P < 0.02) increase in torque per unit area together with a right shift in the IEMG-torque relation and no change in maximum IEMG were observed. Time to peak isometric torque decreased by 45.8% (P < 0.03) but no change in time to maximum IEMG was observed. In conclusion, strength training of the quadriceps results in a variable hypertrophy of its components without affecting its angle-torque relation. The increase in torque per unit area, in the absence of changes in IEMG, may indicate changes in muscle architecture. An increase in muscle-tendon stiffness may account for the decrease in time to peak torque.