Human-robot collaboration is fast becoming a preferable alternative to traditional manual assembly work in manufacturing. Large industrial robots and small force-limited 'co-bots' can now be deployed with sophisticated safety systems to enable human skills and robot skills to be employed most effectively. However, an understanding of the affect this introduction may have on cognitive workload is required prior to their full installation into industry. This paper describes an investigation which explored the effect of working in proximity to a robot on people's workload. The NASA-Task Load Index (NASA-TLX) was used to assess subjective workload for participants after completing trials. Results showed centralised workload levels which were higher for the assembly activity requiring greater fine motor control.