BACKGROUND: Given the global shortage of N95 filtering facepiece respirators (FFP2 in Europe) during the COVID-19 pandemic, KN95 masks (Chinese equivalent of the N95 and FFP2) were imported and distributed in South Africa (SA). However, there are hardly any published independent safety data on KN95 masks. OBJECTIVES: To evaluate the seal, fit and filtration efficiency of several brands of KN95 masks marketed for widespread use in SA healthcare facilities, using standardised testing protocols. METHODS: The verifiability of manufacturer and technical details was first ascertained, followed by evaluation of the number of layers comprising the mask material. The testing protocol involved a directly observed positive and negative pressure user seal check, which if passed was followed by qualitative fit testing (sodium saccharin) in healthy laboratory or healthcare workers. Quantitative fit testing (3M) was used to validate the qualitative fit testing method. The filtration efficacy and integrity of the mask filter material were evaluated using a particle counter-based testing rig utilising aerosolised saline (expressed as filtration efficacy of 0.3 µm particles). Halyard FLUIDSHIELD 3 N95 and 3M 1860 N95 masks were used as controls. RESULTS: Twelve KN95 mask brands (total of 36 masks) were evaluated in 7 participants. The mask type and manufacturing details were printed on only 2/12 brands (17%) as per National Institute of Occupational Safety and Health and European Union regulatory requirements. There was considerable variability in the number of KN95 mask layers (between 3 and 6 layers in the 12 brands evaluated). The seal check pass rate was significantly lower in KN95 compared with N95 masks (1/36 (3%) v. 12/12 (100%); p<0.0001). Modification of the KN95 ear-loop tension using head straps or staples, or improving the facial seal using Micropore 3M tape, enhanced seal test performance in 15/36 KN95 masks evaluated (42%). However, none of these 15 passed downstream qualitative fit testing compared with the control N95 masks (0/15 v. 12/12; p<0.0001). Only 4/8 (50%) of the KN95 brands tested passed the minimum filtration requirements for an N95 mask (suboptimal KN95 filtration efficacy varied from 12% to 78%, compared with 56% for a surgical mask and >99% for the N95 masks at the 0.3 µm particle size). CONCLUSIONS: The KN95 masks tested failed the stipulated safety thresholds associated with protection of healthcare workers against airborne pathogens such as SARS-CoV-2. These preliminary data have implications for the regulation of masks and their distribution to healthcare workers and facilities in SA.