Clouds exert substantial impacts on the global radiation budget and hydrological cycle. We propose a new proxy for the observed low-level cloud amount (LCA), so-called ELF (estimated low-level cloud fraction), and we show that ELF achieves the unprecedented best performance in diagnosing spatiotemporal variations of the observed LCA. Our study can be used to evaluate the performance of GCMs, identify the source of inaccurate simulation of LCA, and better understand climate sensitivity.