Summary Standing positions are extensively used in North American workplaces, whereas in other parts of the world, work is usually performed from a sitting position. Prolonged work while standing has been linked to various symptoms such as low back pain and generalized fatigue. Recent studies have shown a link between low back pain during prolonged standing and coactivation patterns (coordinated activation of muscle pairs) in the lumbo-pelvic muscles. However, these patterns have never been measured for other working positions, or when standing is accompanied by repetitive movement in the upper limbs. The primary objective of this project was to assess the effect of working position (standing, sitting, half-sitting) on lumbo-pelvic muscle coactivation patterns during a repetitive manual task performed in two different environments (laboratory and workplace). A secondary objective was to measure, for each position, the link between muscle patterns and symptoms of discomfort. The working hypothesis was that there would be a link between the coactivation measurements and the manifestation of discomfort symptoms, as already shown for the static standing position. The second hypothesis formulated was that, compared with the literature data showing an increase in lumbo-pelvic coactivation during a task performed standing with no arm movement, the presence of such movements in our study would help attenuate the increase in lumbar coactivation normally observed over time. Lastly, a third hypothesis was that changes in position would influence the lumbo-pelvic muscle coactivation patterns in all the participants, but to a lesser extent in experienced workers than in the novices studied in the laboratory. A group of 11 workers, whose tasks included standing work with repetitive manual movements, followed the research protocol directly in the workplace (5 men, 6 women, average age: 43.2 ± 8.8, years of service: 11.6 ± 9.3). A group of 18 novices, age- and sex- matched, followed the research protocol in the laboratory (10 men, 8 women, average age: 32.4 ± 8.2). During three experimental sessions (presented in random order), the participants in both groups performed a repetitive manual task for half an hour while standing, sitting or half-sitting. The seat used for the sitting position and the stool used for the half-sitting position were adjusted to ergonomic standards. The bilateral lumbo-pelvic muscle activation patterns (erector spinae, external obliques, abdominals, gluteus medius) were measured by surface electrodes, while the discomfort scores were recorded with the help of body diagrams and scales during the manual task and also during the static position intervals. Coactivation measurement was performed according to two methods described in the literature: cross-correlation and mutual information calculation. Variance analyses with the variables Time and Position were applied to the discomfort indices, and the coactivation measurement was performed using two statistical models (one for the laboratory data and another for the workplace data). Analysis of the workplace data showed little effect from position; however, greater bilateral coactivation of the gluteus medius during standing work confirms the results published earlier for tasks performed in a static standing position. The coactivation indices showed no significant change over time, and few of the workers reported discomfort while working in any of the three positions. However, contrary to expectations, the laboratory results showed that several coactivation indices, as well as the number of participants reporting discomfort, were higher for the task performed sitting. In addition, two of the coactivation indices were observed to increase over time. The discrepancy between the workplace results and the laboratory results could possibly be explained by the differences in experience but also in the way the arm movements were performed. The results presented in this report, as well as in other recent results from our laboratory, highlight the impact of working position, experience/expertise and arm movements on the lumbar coactivation indicators, and could thus have an impact on the prediction of the appearance of symptoms connected with manual work in a prolonged position.