At 36 locations worldwide, we estimate the cloud radiative effect (CREatm) on atmospheric solar absorption (ASRatm) by combining ground-based measurements of surface solar radiation (SSR) with collocated satellite-derived surface albedo and top-of-atmosphere net irradiance under both all-sky and clear-sky conditions. To derive continuous clear-sky SSR from Baseline Surface Radiation Network (BSRN) in-situ measurements of global and diffuse SSR, we make use of the Long and Ackerman (2000) algorithm that identifies clear-sky measurements and empirically fits diurnal clear-sky irradiance functions using the cosine of the solar zenith angle as the independent variable. The 11-year average (2000-2010) CREatm (all-sky minus clear-sky) is overall positive at around +11 Wm−2 using direct measurements form ground and space, and at 4 Wm−2 in the CERES EBAF dataset. This discrepancy arises from a potential overestimation in clear-sky absorption by the satellite product or underestimation by the combined BSRN/CERES dataset. The forcing ratio R shows that clouds enhance ASRatm most distinctly at desert-like locations that overall experience little occurrence of clouds. This relationship is captured by both the combined dataset and CERES EBAF.