This study aims to investigate effect of external load and motion repetitiveness on perceived discomfort. An experiment was performed for measuring discomfort scores at varying conditions, in which external load, motion repetitiveness and arm posture were employed as experimental variables. The arm posture was controlled by shoulder flexion and abduction, and by elbow flexion. Fifteen healthy college-age students without history of musculoskeletal disorders voluntarily participated in the experiment. The results showed that the effect of external load, motion repetitiveness and shoulder posture on discomfort were statistically significant, but that elbow posture did not significantly affect discomfort ratings. The effect of external load was much larger than that of any other variables, and that of repetitiveness was second only to external load. Discomfort scores significantly increased linearly as the levels of external load and motion repetitiveness increased. This implies that although they were not fully reflected in the existing posture classification scheme such as OWAS, RULA, etc., the effect of external load and motion repetitiveness should be taken into consideration for precisely quantifying work load in industry. Based on regression analysis, equivalent values of external load and motion repetitiveness in terms of discomfort scores were provided, which would be useful for better understanding the degree of their effect on work load.