The aim was to investigate if the method of normalizing surface electromyography (sEMG) can change results on sex differences in the muscular activation of the shoulder girdle muscles during a simulated work task. sEMG was recorded in 36 asymptomatic participants (17 females, 19 males) from four parts of trapezius and from serratus anterior muscles during a simulated work task. Four normalization methods, one involving maximal voluntary contractions (MVCs) and three involving submaximal voluntary contractions were applied. Sex differences in absolute and normalized amplitude of sEMG were analyzed. The normalization method had a significant influence on the observed sex differences. Females only showed higher sEMG amplitude than males when the sEMGs were normalized to MVC and to a submaximal contraction based on 20% of MVC for the upper trapezius (acromial fibers). Researchers and practitioners should be aware of the impact of the sEMG normalization method in sex differences investigation.