To analyse the muscle adaptations induced by two protocols of isometric training performed at different muscle lengths.

Twenty-eight subjects were divided into three groups: one (K90) performed isometric training of the knee extensors at long muscle lengths (90° of knee flexion) for 8 weeks, and the second group (K50) at short muscle lengths (50°). The subjects of the third group acted as controls. Isokinetic dynamometry was utilized to analyse the net moment–angle relationship and vastus lateralis muscle thickness at three different locations, and pennation angles and fascicle length at 50 % of thigh length were measured at rest with ultrasonography.

Only subjects from K90 group showed significant increases in isokinetic strength (23.5 %, P < 0.001), while K50 group showed no increases in isokinetic strength: (10 %, P > 0.05). There was a shift in the angle of peak torque of the K90 group to longer muscle lengths (+14.6 %, P = 0.002) with greater increases in isokinetic strength, while the K50 angle shifted to shorter muscle lengths (−7.3 %, P = 0.039). Both training groups showed significant increases in muscle thickness, (K90 9–14 % vs. K50 5–9 %) but only K90 significantly increased their pennation angles (11.7 %, P = 0.038). Fascicle lengths remained unchanged.

Isometric training at specific knee angles led to significant shifts of peak torque in the direction of the training muscle lengths. The greater strength gains and the architectural changes with training at long muscle lengths probably come from a combination of different factors, such as the different mechanical stresses placed upon the muscle–tendon complex.