The phase A study of a mid infrared imager and spectrograph for the European Extremely Large Telescope (E-ELT), called METIS, was endorsed in May 2008.1 Two key science drivers of METIS are: a) direct thermal imaging of exo-planets and b) characterization of circumstellar discs from the early proto-planetary to the late debris phase. Observations in the 10μm atmospheric window (N band) require a contrast ratio between stellar light and emitted photons from the exo-planet or the disc of ~" 105. At shorter wavelengths the contrast between star and rejected light from the planet-disc system exceeds ≳"107 posing technical challenges. By means of endto- end detailed simulations we demonstrate that the superb spatial resolution of a 42m telescope in combination with stellar light rejection methods such as coronagraphic or di.erential imaging will allow detections at 10μm for a solar type system down to a star-planet separation of 0.1- and a mass limit for irradiated planets of 1 Jupiter (MJ) mass. In case of self-luminous planets observations are possible further out e.g. at the separation limit of JWST of ~" 0.7-, METIS will detect planets ≳"5MJ. This allows to derive a census of all such exo-planets by means of thermal imaging in a volume limited sample of up to 6pc. In addition, METIS will provide the possibility to study the chemical composition of atmospheres of exo-planets using spectroscopy at moderato spectral resolution (ë/ë ~" 100) for the brightest targets. Based on detailed performance and sensitivity estimates, we demonstrate that a mid-infrared instrument on an ELT is perfectly suited to observe gravitationally created structures such as gaps in proto- and post- planetary discs, in a complementary way to space missions (e.g.JWST, SOFIA) and ALMA which can only probe the cold dust emission further out.