We revisit the environmental Kuznets curve (EKC) hypothesis using 1987-1995 data for Chinese provinces. A comparison of off-sample (1996-2004) predictions to actual emissions indicates that more stringent rules are still needed to fight industrial (waste water and dust) pollution. Auxiliary regressions show that conditional on income,
northern provinces have lower industrial waste water pollution; non-coastal and provinces with smaller secondary industry shares have lower industrial (waste water, COD, and dust) pollution; provinces with smaller state-owned enterprises share have lower industrial COD pollution; and, commitment to control industrial dust pollution
is correlated with local governments’ budget balance.