Abstract.For detection of massive close-in exo-planets around other stars, a contrast of ∼ 0.3% is sufficient and can be achieved with state-of-the-art single-mode fibers as spatial filter. For lower-mass planets, much higher accuracy is neccessary. Apart from the atmosphere, there are several sources of noise which degrade the phase stability of interferometers based on bulk optics by changes in the optical path. For direct detection of low-mass exo-planets and for direct observation of convection on giant stars, our two main science goals, as well as for secondary science goals such as lower-mass faint circumstellar disks around both nearby low-mass stars as well as more distant high-mass stars and for the direct detection of close spectroscopic binary (T Tauri) stars and to follow their orbits for direct dynamical mass determination, we need more than two apertures to fill the uv-plane, again very complicated with bulk-optics. Integrated optics allows to solve these problems and to minimize the components. We have experiences in the development and fabrication of integrated-optical circuits and modulators in various wavelength ranges, i.e. from visible to the telecommunication band, as well as integrated-optical sensor systems like interferometers. Starting from two channel interferometers for the J-band at 1.55µm based on Lithiumniobat, we will develop multichannel-interferometer systems for longer wavelengths. Such systems with new materials and fibers for the near-and thermal infrared can then be used, e.g., for 2nd Gen VLTI instrumentation.