The sensitivity of an Imaging Atmospheric Cherenkov telescope is calibrated by shining, from a distant pulsed monochromatic light source, a defined photon flux onto the mirror. The light pulse is captured and reconstructed by the telescope in an identical fashion as real Cherenkov light. The intensity of the calibration light pulse is monitored via a calibrated sensor at the telescope; in order to account for the lower sensitivity of this sensor compared to the Cherenkov telescope, an attenuator is inserted in the light source between the measurements with the calibrated sensor, and with the telescope. The resulting telescope sensitivities have errors of 10%, and compare well with other estimates of the sensitivity.Imaging Atmospheric Cherenkov Telescopes (IACTs) have evolved into the most powerful tool for the study of galactic and extragalactic γ-ray sources in the TeV energy range [1]. In IACTs, a (frequently tesselated) reflector with areas between a few m 2 and almost 100 m 2 is used to image the Cherenkov photons emitted by an air shower onto a camera consisting of photomultiplier (PM) pixels. The elliptical shower image traces the longitudinal development of an air shower. The long axis of the image points to the image of the source. The shape of the image allows to distinguish, to a certain degree, compact γ-ray induced showers from the more diffuse cosmic-ray showers [2]. The power of IACTs can be improved significantly by operating multiple IACTs in a stereo mode, observing the same shower with several IACTs in coincidence. Stereo imaging allows the unambiguous spatial reconstruction of the direction of individual air showers with a precision of 0.1• and better [3,4,5], and therefore provides the best angular resolution of all tools in γ-ray astronomy.As the field matures, emphasis is shifting from the simple detection of TeV γ-ray sources to the precise determination of source fluxes and their spectra. The spectra contain important clues both concerning the acceleration mechanisms