AER is developing a testbed capable of simulating lidar sensors, such as those used for Differential Absorption LIDAR (“DIAL”) measurements. The testbed is designed for maximum flexibility to allow for the rapid prototype and development of trace gas remote sensing systems. The input atmospheric profile set (used to specify the temperature, moisture and molecular constituent profiles) is constructed from actual numerical weather prediction (NWP) model fields drawn from a global database of model outputs. The testbed has the capability of perturbing these input model fields to simulate errors in the specification of the atmospheric state or to provide a set of test cases spanning a wide range of conditions with realistic, but highly variable, structure. The profiles are used in conjunction with automated LBLRTM runs for line-by-line optical depth calculations over the desired spectral region. LBLRTM can also be coupled with the CHARTS multiple scattering code for cases requiring multiple scattering calculations (aerosols and clouds). The input parameters include sensor altitude, viewing geometry, surface altitude and reflectivity, and cloud/aerosol parameters (such as density, location and specific phase functions). Additional inputs included in the testbed are sensor-related parameters such as the laser line shape, detector noise and laser wavelength stability. The testbed may be coupled to various geophysical parameter retrieval algorithms, including those specific to the DIAL technique. The geophysical parameter retrievals can then be computed and compared with the truth. This paper will describe the testbed architecture and the process by which the input NWP profiles are generated, and will provide example LIDAR simulations for trace gas retrievals with a DIAL system.