The employment impact of environmental policies is an important question for policy makers. We examine the effect of increasing the stringency of environmental policy across a broad set of policies on firms’ labor demand, in a novel identification approach using Worldscope data from 31 countries on firm-level CO2 emissions. Drawing on evidence from as many as 5300 firms over 15 years and the OECD environmental policy stringency (EPS) index, it finds that high emission-intensity firms reduce labor demand upon impact as EPS is tightened, whereas low emission-intensity firms increase labor demand, indicating a reallocation of employment. Moreover, tightening EPS during economic contractions appears to have a positive effect on employment, other things equal. Quantifications exercises show modest positive net changes in employment for market-based policies, and modest negative net changes for non-market policies (mainly emission quantity regulations) and for the combined aggregate EPS. Within market-based policies, the percent decline in employment in high-emission firms (correspondingly the increase in low-emission firms) for a unit change in a policy index is smallest (largest) for trading schemes (“green” certificates, and “white” certificates)—although stringency is not comparable across indices. Finally, the employment effects of EPS are not persistent.