The amount of longwave radiation reaching the surface is a necessary parameter for climate modeling. As no extensive array of surface-based stations exists for monitoring the downwelling longwave surface flux (F↓(0)), satellites are expected to provide the data necessary for its computation. Before global fields of satellite-derived values of F↓(0) can be used for climate modeling, computations of F↓(0) need to be validated with surface measurements. In this study the validation of computations of F↓(0) is carried out for both clear and cloudy conditions using data from the 1989 phase of the First ISLSCP Field Experiment (FIFE 1989). FIFE 89 data is particularly useful for this study, as it contains lidar measurements of cloud base height and cloud fraction, thereby eliminating the need to estimate these parameters from other available data. Using a combination of the delta-Eddington approximation for atmospheric scattering and the LOWTRAN 7 transmission functions, the computation of F↓(0) was carried out with particular emphasis on the reliability of the available data. Our results show that the two largest sources of discrepancies between the measured and calculated F↓(0) values are the lack of knowledge of cloud optical depth and the measurement error inherent in pyrgeometer instruments. The uncertainty due to the lack of optical depth information ranges from 1 to 22 W/m2, while the uncertainty of the pyrgeometer measurements ranges from 15 to 23 W/m2. For both clear and cloudy conditions the radiative transfer model was able to predict surface irradiance values within 4% of the measured values. In each case the standard deviation of the calculated values fell within the accuracy range of the pyrgeometer. As more detailed observations become available, such as those from the Atmospheric Radiation Measurement program, we anticipate a refinement in our ability to compute F↓(0). ¿ American Geophysical Union 1995