Assembly lines are key manufacturing systems, but they often face issues like operation time variability or ergonomic risks for workers. These challenges lead to operational inefficiencies, including a rise in musculoskeletal disorders (MSDs), and increased economic costs due to absenteeism and worker compensation. This work aims to help make the workplace more efficient and cost-effective by improving worker conditions. Moreover, in addition to the operational difficulties involved, the variability also complicates the accurate assessment of a worker's physical strain. We introduce a new strategy for scenario-based optimization of the assembly line balancing problem, modeling ergonomics with a fatigue and recovery criterion. Here, operation times can vary according to different scenarios, each associated with a probability of occurrence. Our objective is to minimize the fatigue level of workers for a given percentage of scenarios. To achieve this goal, we propose an Integer Linear Programming (ILP) approach that prioritizes worker health under fixed economic conditions. An Iterative Dichotomic Search algorithm is used to solve the problem, and a numerical example is presented. Our method is a new approach that uses variable operation times to better align with real-world assembly line uncertainties. To the best of our knowledge, our model is the first to deal with ergonomics in cases with variable operation times for manual workers, thus enhancing its practical applicability and effectiveness in real-world industrial settings. Practitioners benefit from our model’s adaptability to operation variations, offering practical solutions for decision-making and enhanced operational management.