The purpose of this study was to evaluate the ergonomic risk factors associated with manual lifting tasks using surface electromyography (EMG). 13 volunteers lifted loads of 6 and 13 kg at two speeds and at two horizontal distances in 3 different postures and three boxes of different sizes, from floor to knuckle height, performing 72 lifting tasks. For each lift, the surface electromyography signals from the erector spinae muscles, bilaterally at T10 and L3, was recorded. The ergonomic risk factors associated with manual lifting tasks were evaluated by comparing the average amplitude of EMG signals from the erector spinae muscles. The EMG average amplitude for lifting the load of 13 kg was 14.3% greater than that for lifting the load of 6 kg (t=-10.93, P<0.01). The EMG average amplitude at the site of L3 was 10.3% greater than that at the site of T10 (t=-7.98, P<0.01). The EMG average amplitude when performing "fast" lift was 5.9% greater than the "slow" lift (t=-4.63, P<0.01). The posture of lifting affected the EMG average amplitude. It was lowest with semi-squat posture and greatest with squat posture (F=27.76, P<0.01). The result of multiple stepwise regression analysis showed that the loads of lifting, the size of box, horizontal distance, posture of lifting, the site of the spine subjected to force, lifting speed were the factors affecting the EMG average amplitude. The most significant factor was the loads of lifting, followed by the site of the spine subjected to force and the lifting speed in terms of risk. The ergonomic risk factors associated with manual lifting tasks includes the loads, posture, lifting speed, horizontal distance, the site of the spine subjected to force etc. The results of signal amplitude of EMG from the erector spinae muscles showed that semi-squat posture is the best posture for lifting tasks.