Muscle quality is defined as strength per unit muscle mass. The aim of this study was to measure the maximal voluntary isometric torque of the knee extensor and flexor muscle groups in healthy older women and to develop an index of muscle quality based on the combined knee extensor and flexor torque per unit lean tissue mass (LTM) of the upper leg. One hundred and thirty‐six healthy 50‐ to 70‐year‐old women completed an initial measurement of isometric peak torque of the knee extensors and flexors (Con‐Trex MJ; CMV AG, Dubendorf, Switzerland) that was repeated 7 days later. Subsequently, 131 women returned for whole‐ and regional‐body composition analysis (iDXA™; GE Healthcare, Chalfont St Giles, Buckinghamshire, UK). Isometric peak torque demonstrated excellent within‐assessment reliability for both the knee extensors and flexors (ICC range: 0·991–1·000). Test–retest reliability was lower (ICC range: 0·777–0·828) with an observed mean increase of 5% in peak torque [6·2 (17·2) N m] on the second day of assessment (P<0·001). The relative mean decrease in combined isometric peak torque (−12·2%; P = 0·001) was double that of the relative, non‐significant, median difference in upper leg LTM (−5·3%; P = 0·102) between those in the 5th and 6th decade. The majority of difference in peak isometric torque came from the knee extensors (15·1 N m, P<0·001 versus 2·4 N m, P = 0·234). Isometric peak torque normalized for upper leg LTM (muscle quality) was 8% lower between decades (P = 0·029). These findings suggest strength per unit tissue may provide a better indication of age‐related differences in muscle quality prior to change in LTM.