This study investigated the effects of initial load height and foot placement instruction in four lifting techniques: free, stoop (bending the back), squat (bending the knees) and a modified squat technique (bending the knees and rotating them outward). A 2D dynamic linked segment model was combined with an EMG assisted trunk muscle model to quantify kinematics and low back loading in 10 subjects performing 19 different lifting movements, using 10.5 kg boxes without handles. When lifting from a 0.05 m height with the feet behind the box, squat lifting resulted in 19.9% (SD 8.7%) higher net moments (p < 0.001) and 17.0% (SD 13.2%) higher compression forces (p < 0.01) than stoop lifting. This effect was reduced to 12.8% (SD 10.7%) for moments and a non-significant 7.4% (SD 16.0%) for compression forces when lifting with the feet beside the box and it disappeared when lifting from 0.5 m height. Differences between squat and stoop lifts, as well as the interaction with lifting height, could to a large extent be explained by changes in the horizontal L5/S1 intervertebral joint position relative to the load, the upper body acceleration, and lumbar flexion. Rotating the knees outward during squat lifts resulted in moments and compression forces that were smaller than in squat lifting but larger than in stoop lifting. Shear forces were small ( < 300 N) at the L4/L5 joint and substantial (1100-1400 N) but unaffected by lifting technique at the L5 /S1 joint. The present results show that the effects of lifting technique on low back loading depend on the task context. © 2004 Taylor & Francis Ltd.