Investigating the effects of workload on body kinematics is the first step to identify, monitor, and ultimately reduce the incidence of fatigue, a prevalent phenomenon in the workplace that leads to chronic disorders, loss of productivity, and absenteeism (Lu, Megahed, Sesek, & Cavuoto, 2017, In Press; Ricci, Chee, Lorandeau, & Berger, 2007). In fa- tigue monitoring, kinematic measures including acceleration, jerk, and body posture have been found to be informative (Lu et al., 2017, In Press; Maman, Yazdi, Cavuoto, & Megahed, 2017; Ricci et al., 2007); however, none of the previous studies have considered a comprehensive set of these kinematic metrics during simulated manufacturing tasks. This study assessed the effects of duration, ease of task, and age as three factors on different body kinematic metrics and subjective ratings as a substitute “ground truth” for fatigue development. This will serve to inform feature selection for modeling fatigue development over a broad range of industrial tasks. Nineteen participants (divided into two age groups of younger (<25, 6 males and 4 females) and older (>40, 8 males and 1 female)) completed three, three-hour sessions of parts assembly (light), supply pickup and insertion (moderate), and manual material handling (difficult). Inertial measurement units (IMUs) were attached on right wrist, middle of the trunk, the right side of the hip, and right ankle. The mean and peak values of acceleration, jerk, and posture for each body location along with the minimum value (with respect to the horizontal plane) of trunk bending posture were considered as the kinematic variables of interest. The Borg rating of perceived exertion (RPE) and subjective fatigue level (SFL) were recorded at the start of each session, and then every ten minutes for RPE, and every thirty minutes for SFL. Perceived workload, rated using the NASA Task Load Index (TLX), was obtained every one hour. The TLX, RPE, and SFL at the end of hour 1 (the first time point where all three ratings are obtained) were considered as the pre-fatigue values and at hour 3 of the tasks as the post-fatigue. Similarly, the pre-fatigue kinematic data of the IMUs was the period from minutes 10 to 20 and the post-fatigue data was the period from minutes 160 to 170. The results of repeated measures ANOVA showed a significant time ( p < 0.001) effect on all three subjective ratings. In addition, time and age interacted to affect RPE ( p = 0.018) with a 36% increase in younger and 28% in the older group. Time had significant effects only on a few of kinematic variables including mean trunk acceleration (8% decreased), mean trunk posture (3% less bent), and peak hip acceleration (10% increased) after fatigue. Moreover, there was a significant age and task interaction for peak hip acceleration (~1 m/s2 decreased), mean and peak leg posture (~4o increased and ~12o decreased, respectively), and minimum trunk posture (~5o increased) from the younger group to the older group. In addition, there was significant interaction ( p = 0.011) between time and task in bending posture denoted by hip and trunk, which provides insight into the different effects of fatigue on different tasks, i.e., 2% more bending after fatigue in manual material handling and supply pickup and insertion in comparison to parts assembly. This increased bending angle following fatigue was in agreement with the findings of Strohrmann, Harms, Kappeler-Setz, and Troster (2012). There were significant differences between the younger group to the older group in terms of kinematics, i.e., peak hip acceleration, mean and peak leg posture, and minimum trunk posture that may be attributable to different quadriceps strength and postural stability between the age groups. Overall, the results present a set of kinematic parameters influenced by fatigue; however, further analysis is required to explore more temporal and spatial movement variables from IMUs for a better understanding of fatigue effects and indicators.