The dataset contains pressure insole data from twenty subjects who performed five tasks, comprising of two common daily activities (standing and walking), and three industry-focussed tasks (manual handling, assembly and pick and place). The speed and order in which a given task was completed was not prescribed. The data pertains to the areas of human factors, ergonomics and occupational health and safety research, among others, and enables an understanding of the force distributions involved in common tasks as well as physical and manufacturing type tasks. This work is licensed under the Creative Commons Attribution 4.0 International License. If you are using our database to evaluate your methods, please cite: O’Sullivan, P., Menolotto, M., Visentin, A., O’Flynn, B. and Komaris, D.S., 2024. AI-based Task Classification with Pressure Insoles for Occupational Safety. IEEE Access.