This paper attempts to examine if the “strong” version of Porter Hypothesis is supported in China by investigating how different regulatory instruments and the relative stringency impact “green” productivity. We use a slacks-based measure (SBM) and Luenberger Productivity Index, accounting for undesirable outputs, to evaluate the industrial “green” productivity growth rates of China's 30 provinces. The estimates imply an unsustainable development model in China with significant regional differences. By employing a panel threshold model and a province-level panel dataset during 2000–2012, empirical results show that both command-and-control and market-based regulation have a non-linear relationship with and can be positively related to “green” productivity but with different constrains on regulation stringency: there are double thresholds with the command-and-control and exists an optimal range of stringency for productivity improvement; while a single threshold has been found with the market-based regulation and its current stringency is reasonable for most of provinces. Moreover, based on China's reality, the productivity effect driven by market-based regulation is much stronger than that of the command-and-control. The mechanism of informal regulation is much more complicated. Consequently, we find evidence to support the “strong” Porter Hypothesis that reasonable stringency of environmental regulations may enhance rather than lower industrial competitiveness. •Examine the “strong” version of Porter Hypothesis in the case of China•Environmental total factor productivity accounting for undesirable outputs•The command-and-control, market-based and informal environmental regulation•Investigate the non-linear relationship between regulation and “green” productivity•The fixed-effect panel threshold model