It is frequently hypothesized that environmental management systems (EMSs) may improve a firm?s environmental performance. Whether or not this hypothesis is true is as important from the perspective of environmental policy as questions relating to the relevant incentives for (1) a firm?s voluntary adoption of an EMS and (2) its environmental innovation behavior. Based on ample empirical evidence for German manufacturing, this paper addresses these issues on the basis of a recursive bivariate probit model that explicitly takes into account that a facility?s decision on innovation activities is correlated with the decision on EMS certification. Our empirical results indicate that environmental innovation activities are not associated with EMS certification nor any other single policy instrument. Rather, innovation behavior seems to be correlated to the stringency of environmental policy.