Background  Occupational hazards such as dust and toxic chemicals in the production process of automobile manufacturers will adversely affect the health of workers.

Objective This study applies three semi-quantitative risk assessment methods in GBZ/T 298-2017 Guidelines for occupational health risk assessment of chemicals in the workplace to assess the occupational health risk in automobile manufacturing enterprises and explores their applicability.

Methods Judgment sampling method was used to select four out of six automobile manufacturing enterprises in Wuhan City, Hubei Province to conduct occupational health risk assessment for the main positions exposed to occupational hazards, and the results of three occupational health risk assessment methods were compared and verified.

Results The major occupational hazards of the four automobile manufacturers were welding fume, grinding wheel dust, manganese and its inorganic compounds, nitrogen dioxide, toluene, xylene, methyl ethyl ketone, butyl acetate, butanol, and isopropyl alcohol. The results of exposure ratio method showed that the melt inert-gas (MIG) welding positions of companies C and D were high-risk positions. The results of index method showed that the spot welding and polishing positions of the four companies and the MIG welding positions of companies A, C, and D were medium-risk positions. The results of composite index method were the same as the index method. When exposure concentration (EC) was less than 1/2 of the relevant national occupational exposure limits (OELs), the risk index (R) of the exposure ratio method (1.694±0.433) was lower than those of the index method (2.344±0.317) and the composite index method (2.327±0.317) (P < 0.001). When 1/2 OELs ≤ EC < OELs, there was no significant difference in the R values of the exposure ratio method (2.966±0.138), the index method (2.916±0.206), and the composite index method (2.924±0.195) (P>0.05). When EC ≥ OELs, the R of the exposure ratio method (3.398±0.289) was higher than those of the index method (2.802±0.283) and the composite index method (2.887±0.279) (P < 0.001). The consistency of the assessment results between the exposure ratio method and the index method was poor (weighted Kappa=0.118, P < 0.001), that between the exposure ratio method and the composite index method was also poor (weighted Kappa=0.136, P < 0.001), and that between the index method and the composite index method was excellent (weighted Kappa=0.977, P < 0.001). For positions with EC exceeding OELs, the three assessment results were all above medium risk, consistent with the determination results of OELs. The manual spraying positions were evaluated as negligible-risk positions and low-risk positions by the three assessment methods, inconsistent with the results of occupational health examination.

Conclusion The three semi-quantitative risk assessment methods can identify key positions with occupational health hazards, but the results for the manual spraying positions are conservative. Because the results of the three methods are related to EC, it is suggested to choose appropriate assessment methods according to the EC of target occupational hazardous factors.