The assessment of length-tension relationship in human muscle is cumbersome, and is usually derived from isometric measurements of the angle-torque relationship. Slow dynamic contractions may offer an efficient and accurate alternative (Proske et al., 2004). The aim of this study was to compare these two methods for measurement of torque-angle relationship in human knee extensors and flexors. Twenty-two male Gaelic football players (aged 21.9±1.4 yrs, height 184.5±6.0 cm, body mass 85.7±7.7 kg, all data mean±SD) participated in this study after giving written informed consent. Both dynamic concentric and isometric measurements of left and right legs were made in a single testing period one week after a familiarisation session. Volunteers were seated on an isokinetic dynamometer with knee range of motion set from 80° to 160° (where 180° is full extension). For each leg, a single maximal dynamic contraction at an angular velocity of 0.52rad.s^-1 (30^o/sec) was performed for both knee extension and flexion. Volunteers then performed two sets of isometric maximal voluntary contractions of the knee extensors and flexors every 10 degrees through the range of motion with 30 seconds rest between contractions. Angle of peak torque was extracted using both methods and compared using t-tests. Torque-angle relationship were normalised for peak torque in each individual and the two methods were compared using repeated measures ANOVA; Bland-Altman plots were used to assess agreement between the two methods. Isometric peak torque for right knee extension was 311.1±57.1 N.m and 300.6±57.0 N.m measured in concentric contraction. Flexion peak torque was 178.6±28.1 N.m (isometric) and 175.4±31.1 N.m (concentric). Bland-Altman plots indicated that the limits of agreement between the two methods were +82.1 and -61.2 N.m for right knee extension and +50.0 to -43.5 N.m for flexion, suggesting poor agreement between the two methods. Angle of peak torque for right knee extension was 105.9±7.3° when measured isometrically and 111.5±6.8° measured concentrically (significantly different, P<0.01, t-test). The same values for flexion were 157.3±4.6° and 159.5±5.6° respectively. Figure 1 illustrates the normalised torque-angle relationships for knee extension using both methods. There was a significant method-angle interaction (P<0.01) for both knee flexion and extension, suggesting a lack of agreement between the methods. These results suggest that dynamic measurement of angle-torque relationship at this angular velocity does not closely agree with isometric measurement.