Free-space optical (FSO) communication is a very attractive technology offering very high throughput without spectral regulation constraints, yet allowing small antennas (telescopes) and tap-proof communication. However, the transmitted signal has to travel through the atmosphere where it gets influenced by atmospheric turbulence, causing scintillation of the received signal. In addition, climatic effects like fogs, clouds and rain also affect the signal significantly. Moreover, FSO being a line of sight communication requires precise pointing and tracking of the telescopes, which otherwise also causes fading. To achieve error-free transmission, various mitigation techniques like aperture averaging, adaptive optics, transmitter diversity, sophisticated coding and modulation schemes are being investigated and implemented. Evaluating the performance of such systems under controlled conditions is very difficult in field trials since the atmospheric situation constantly changes, and the target scenario (e.g. on aircraft or satellites) is not easily accessible for test purposes. Therefore, with the motivation to be able to test and verify a system under laboratory conditions, DLR has developed a fading testbed that can emulate most realistic channel conditions. The main principle of the fading testbed is to control the input current of a variable optical attenuator such that it attenuates the incoming signal according to the loaded power vector. The sampling frequency and mean power of the vector can be optionally changed according to requirements. This paper provides a brief introduction to software and hardware development of the fading testbed and measurement results showing its accuracy and application scenarios.