Does Thermodynamics Have a Reversibility Problem?

Abstract Does thermodynamics have a reversibility problem? I argue that it does not. Insofar as thermodynamics exhibits an arrow of time, it does so due to a failure of conservation or special initial conditions, which are exactly the circumstances under which statistical mechanics exhibits an arrow of time. The arrows of thermodynamics and of statistical mechanics are therefore aligned, and so pose no obstacle to the reduction of thermodynamics to statistical mechanics.