Shoulder fatigue has been suggested to be a useful risk indicator for shoulder disorders in repetitive, low-force work tasks. In contrast to high-force or purely isometric tasks, it is unclear whether measurable fatigue develops in realistic low-force work. The question addressed in this review was: 'Is there evidence of objective signs of fatigue in the shoulder region in realistic, low-force work tasks?' Studies on objective measures of fatigue applied in realistic low-force work tasks were systematically reviewed, using a task duration of more than 1 h and an intensity level of less than 20% maximum voluntary contraction (MVC) for the median trapezius activation level as inclusion criteria. Thirteen studies were found to fulfil the criteria. All these studies addressed fatigue-related changes in the electromyographic signal in the descending part of the trapezius muscle. Seven did find a combination of frequency decrease and amplitude increase over time, which is generally considered as an objective manifestation of fatigue. Thus, there is evidence of objective signs of fatigue in some of the realistic, low-force tasks. The intensity level appeared to be a main determinant here. In the studies demonstrating signs of fatigue an intensity level of 15%MVC or more was used, while the intensity level in the studies with a negative result was generally lower.