What is the central question of the study? Do changes in neuromuscular activation contribute to the knee extensor angle-torque relationship? What is the main finding and its importance? Both agonist (quadriceps) and antagonist coactivation (hamstrings) differed with knee joint angle during maximal isometric knee extensions and thus both are likely to contribute to the angle-torque relationship. Specifically, two independent measurement techniques showed quadriceps activation to be lower at more extended positions. These effects might influence the capacity for neural changes in response to training and rehabilitation at different knee joint angles. The influence of joint angle on knee extensor neuromuscular activation is unclear, owing in part to the diversity of surface electromyography (sEMG) and/or interpolated twitch technique (ITT) methods used. The aim of the study was to compare neuromuscular activation, using rigorous contemporary sEMG and ITT procedures, during isometric maximal voluntary contractions (iMVCs) of the quadriceps femoris at different knee joint angles and examine whether activation contributes to the angle-torque relationship. Sixteen healthy active men completed two familiarization sessions and two experimental sessions of isometric knee extension and knee flexion contractions. The experimental sessions included the following at each of four joint angles (25, 50, 80 and 106 deg): iMVCs (with and without superimposed evoked doublets); submaximal contractions with superimposed doublets; and evoked twitch and doublet contractions whilst voluntarily passive, and knee flexion iMVC at the same knee joint positions. The absolute quadriceps femoris EMG was normalized to the peak-to-peak amplitude of an evoked maximal M-wave, and the doublet-voluntary torque relationship was used to calculate activation with the ITT. Agonist activation, assessed with both normalized EMG and the ITT, was reduced at the more extended compared with the more flexed positions (25 and 50 versus 80 and 106 deg; P ≤ 0.016), whereas antagonist coactivation was greatest in the most flexed compared with the extended positions (106 versus 25 and 50 deg; P ≤ 0.02). In conclusion, both agonist and antagonist activation differed with knee joint angle during knee extension iMVCs, and thus both are likely to contribute to the knee extensor angle-torque relationship.