Wearing face masks have been implemented as a public and personal health control measure against the spread of coronavirus disease (COVID-19). However, the protection level of nonmedical face masks, such as women face veils, is still uncertain. This study aimed to assess the filtration efficiency (FE; percentage of particles retained by a mask) of different types of medical masks (either as sealed or unsealed, single or doubled), non-medical masks (cloth masks) and face veils. FE of face masks was evaluated using an in-house 3D-printed air duct connected to the Aerotrak particle counter with a capability of counting particle sizes of 0.3, 0.5, 0.7, 1, 2 and 5 μm. A set of 10 earloop surgical masks,10 tie-on surgical masks, 3 triple-layers reusable cloth masks and 3 types (short, medium and long) of traditional face veils were included in the study. The unsealed surgical masks showed intermediate FE (36.54-80.58%), with no observed differences between tie-on and earloop or single and doubled masks. For each mask type, the mean FE values of sealed surgical masks (FE≥99.16%) was significantly higher (P<0.001) than the unsealed ones (FE≤80.58%). No significant difference was observed in the mean FE values between unsealed surgical masks and either cloth masks (FE=23.19-75.35%, P=0.26) or face veils (FE=19.10- 70.68%, P=0.14). However, a mockup experiment showed that wearing a surgical mask under the face veil significantly improve the FE (33.73-79.18%; P<0.001). We conclude that besides sealed surgical masks that ensure optimal filtration under the experimental conditions, the unsealed surgical and cloth masks and face veils showed comparable performance and acceptable protection at 5 μm particle size, which is the most relevant particle size associated with COVID-19 infectious droplets. Wearing a surgical mask under the face veil significantly improves the FE compared to wearing a face veil alone.