Direct capture efficiency of a local exhaust system is defined by introducing an imaginary control box surrounding the contaminant source and the exhaust opening. The imaginary box makes it possible to distinguish between contaminants directly captured and those that escape. Two methods for estimation of direct capture efficiency are given: (I) a numerical method based on the time-averaged Navier-Stokes equations for turbulent flows; and (2) a field method based on a representative background concentration. Direct capture efficiency is sensitive to the size of the control box, whereas its location is less important for the case studied. The choice of sampling strategy to obtain a representative background concentration is essential as substantial differences on direct capture efficiency are found . Recommendations are given.