The purposes of this study were to examine the effect of two different pelvic alignments and the Valsalva maneuver on electromyographic (EMG) activity of the erector spinae during squat lifting and lowering, and to find an efficient method for squat lifting and lowering. Twenty hea1thy men in their twenties lifted and lowered loads using four different methods: 1) anterior pelvic tilt position with the Valsalva maneuver, 2) anterior pelvic tilt position without the Valsalva maneuver, 3) posterior pelvic tilt with the Valsalva maneuver, 4) posterior pelvic tilt without the Valsalva maneuver. The EMG activity of erector spinae was recorded during both lifting and lowering with each method. The EMG activity of each individual was normalized to EMG activity produced by muscle during maximal voluntary contraction. Two-way analysis of variance for repeated measures ($2{ imes}2$) was used to analyze the effect of the two factors: 1) pelvic tilt position (anterior pelvic tilt, posterior pelvic tilt), 2) the Valsalva maneuver (with and without). Analysis was performed separately for the lifting and lowering. The results were as follows: 1) EMG activity of erector spinae was greater when the pelvis was tilted anteriorly than when the pelvis was tilted posteriorly during squat lifting and squat lowering. 2) There was no difference between EMG activity of erector spinae with the Valsalva maneuver and EMG activity of erector spinae without the Valsalva maneuver during squat lifting and squat lowering. These results suggest that the greater EMG activity of erector spinae with an anterior pelvic tilt position during squat lifting and squat lowering may ensure optimal muscular support for the spine while handling loads, but the Valsalva maneuver may have less effect on erector spinae.