The modelling and enforcement of typestate constraints in object oriented languages has the potential to eliminate a variety of common and difficult to diagnose errors. While the theoretical foundations of typestate are well established in the literature, less attention has been paid to the practical aspects: is the additional complexity justifiable? Can typestate be reasoned about effectively by "real" programmers? To what extent can typestate constraints be inferred, to reduce the burden of large type annotations? This thesis aims to answer these questions and provide a holistic treatment of the subject, with original contributions to both the theorical and practical aspects of typestate.