In local exhaust ventilation systems, external turbulence from e.g. human movements can affect the capture efficiency of the system considerably. In this study, experimental and numerical (CFD) approaches were utilized to evaluate the effect of human movement on the capture efficiency of a local exhaust ventilation system with an exterior circular hood. Human movements were simulated by back-and-forth movements of three human-sized moving objects: a flat plate (CFD + experimental), a cylinder (CFD) and a human-shaped manikin (CFD), respectively. The experiments consisted of tracer gas measurements in a full-scale test room. The numerical simulations included dynamic mesh methods to handle object movements. The results showed reasonable agreement between numerical and experimental results regarding the capture efficiency at different movement frequencies and exhaust flow rates, indicating that CFD is a feasible method for investigating complex flows of the studied kind. In comparison with the moving manikin, the moving plate caused significantly lower capture efficiency, whereas the moving cylinder yielded higher values. Overall, the results with the cylinder as moving object proved more similar to those of the manikin than the results with the flat plate. These findings have particular relevance towards existing test standards that stipulate the use of a moving flat plate in similar test situations. Further, some parameter variations verified that local exhaust capture efficiency increases by increasing the exhaust air flow rate and movement time interval, and also by decreasing the distance between contaminant source and exhaust hood opening. Also increasing the distance between the movable object and the contaminant source, as well as decreasing the diameter of the exhaust hood opening increased the capture efficiency.