After reviewing the strong CP problem and its solution via the QCD axion, I discuss the experimental status of axion searches. A major goal of axion searches is reaching inside the parameter space region of realistic axion models. Currently, the boundaries of this region depend on somewhat arbitrary criteria, and it would be highly desirable to specify them in terms of precise phenomenological requirements. In this talk I put forth a definition for a phenomenologically preferred axion window as the region encompassing all hadronic axion models which i) do not contain cosmologically-stable strongly-interacting relics and ii) do not induce Landau poles below the Planck scale.