Observations of clouds from the ground-based U.S. Department of Energy Atmospheric Radiation Measurement (ARM) program and satellite-based A-train are used to compute cloud radiative forcing profiles over the ARM Darwin, Australia site. Cloud properties are obtained from both radar (the ARM Millimeter Cloud Radar (MMCR) and the CloudSat satellite in the A-train) and lidar (the ARM Micropulse lidar (MPL) and the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) satellite in the A-train) observations. Cloud microphysical properties are taken from combined radar and lidar retrievals for ice clouds and radar-only or lidar-only retrievals for liquid clouds. Large, statistically significant differences of up to 1.43 K/d exist between the mean ARM and A-train net cloud radiative forcing profiles. The majority of the difference in cloud radiative forcing profiles is shown to be due to a large difference in the cloud fraction above 12 km. Above this altitude, the A-train cloud fraction is significantly larger because many more clouds are detected by CALIPSO than by the ground-based MPL. It is shown that the MPL is unable to observe as many high clouds as CALIPSO due to being more frequently attenuated and a poorer sensitivity. We also isolate the difference in cloud radiative forcing due to sampling and retrieval differences which are of comparable importance but are of smaller impact than cloud fraction differences. This study demonstrates that A-train observations are better suited for the calculation of cloud radiative forcing profiles at Darwin. In addition, we find that it is necessary to supplement CloudSat with CALIPSO observations to obtain accurate cloud radiative forcing profiles.