PurposeThe study aims to explore macroeconomic and banking industry-specific determinants of non-performing loans (NPLs) for Chinese banks, spanning from 2005 to 2014.Design/methodology/approachIt uses three different models to explore the determinants. The first model has only macroeconomic variables as regressors; the second model has only banking industry-specific variables as independent variables; and the third model has macroeconomic and banking industry-specific variables as explanatory variables. Furthermore, system generalized method of moments estimation technique has been used to measure the coefficients of independent variables.FindingsGross domestic product (GDP) growth rate, effective interest rate, inflation rate, foreign exchange rate, type of bank, bank risk-taking behavior, ownership concentration, leverage and credit quality are significant determinants of NPLs in Chinese banks. Furthermore, the determinants of NPLs for listed and unlisted banks differ. Determinants of NPLs of listed banks include GDP, bank risk-taking behavior and credit quality. However, variation in NPLs of unlisted banks is explained by GDP, inflation rate, foreign exchange rate, bank risk-taking behavior, leverage and credit quality.Originality/valueThis study also finds that using only macroeconomic or banking industry-specific variables as regressors is not a right approach because it may lead to erroneous conclusions.