Weakly Interacting Massive Particle (WIMP) direct detection experiments are just reaching the sensitivity required to detect Galactic dark matter in the form of neutralinos (or indeed any stable weakly interacting particle). Detection strategies and data analyses are often based on the simplifying assumption of a standard spherical, isothermal halo model, but observations and numerical simulations indicate that galaxy halos are in fact triaxial and anisotropic, and contain substructure. the annual modulation and direction dependence of the event rate (due to the motion of the Earth) provide the best prospects of distinguishing WIMP scattering from background events, however these signals depend sensitively on the local WIMP velocity distribution. I briefly review the status of WIMP direct detection experiments before discussing the dependence of the annual modulation signal on astrophysical input, in particular the structure of the Milky Way halo, and the possibility that the local WIMP distribution is not smooth.