Musculoskeletal stress markers (MSM) have been used to reconstruct activity patterns and labour intensity of past populations. Age and size have been found to correlate with MSM, and these aspects should be considered in activity reconstructions. The aim of this study was to find out the nature and the effects of labour intensity, age and size on MSM. Study material were skeletons (N = 108) of individuals of known age, sex and occupation housed at the Natural History Museum, Finland. MSM were scored for Pectoralis major, Deltoid, Teres major and Biceps brachii. These scores were combined to reflect total activity of an individual. Geometric mean of humeral measurements was used as a size indicator and radial tuberosity size was used as a muscle size indicator. Factors explaining MSM were studied using ANCOVA. This included age, size, muscle size, sex, labour intensity, and their interactions. Age and muscle size were the most significant factors explaining MSM, where muscle attachment areas and MSM grow with advancing age. Muscle attachment areas and skeletal frame size were also found to correlate. Least squares regression parameter estimates were used to study the effects of labour intensity, sex and side on MSM. It was found that in early life scores are higher in heavy labour group, but there is less age-related increase in these scores. This could mean that bone is unable to respond to heavy and continuous loading with surface structure. Therefore labour intensity cannot be reliably recorded in old individuals. Also age and size (as reflected in muscle attachment area) affect MSM and these aspects should be considered before making assumptions on labour intensity. Copyright © 2009 John Wiley & Sons, Ltd.