Results from extensive microwave coupling measurements on a generic missile are presented. Angular resolved plane wave coupling to internal probes and coupling in the statistically isotropic environment of a reverberation chamber are analysed. It is shown that theoretical expressions from antenna theory, using the electrical size of the object, can be used to determine the angular step angle necessary to resolve the receiving pattern. The average probe receiving cross sections measured in the two chambers show a fairly good agreement, which is expected given the statistically isotropic environment in the reverberation chamber. The apparent directivity, i.e. the difference between the maximum measured plane wave cross section and the average, is around 10 dB. The apparent directivity is fairly well predicted by theoretical expressions derived by NIST (National Institute of Standards and Technology, USA). It is shown that the frequency step needed to resolve frequency variations is given by the Q-bandwidth of the cavity. A statistical analysis of the coupling data shows that the variations as function of angle of incidence and polarisation in the anechoic chamber, is not statistically equivalent to the variation as function of stirrer position in the reverberation chamber. This illustrates the difficulty to correlate the two types of environment. An FDTD simulation shows a very good match to measured data for the strong coupling case. The results of this study, supported by similar investigations of several other objects, provide a fundamental guideline on how to perform and evaluate immunity testing of real systems and especially on how to relate testing in anechoic chamber to testing in reverberation chamber.