This paper considers optimal monetary policy in the context of the central bank adopting an asymmetric objective function. Our results show that under asymmetric preferences, many of the extant results on the time consistency problem need no longer hold.