A new diagnostic metric based on soil moisture bimodality is developed in order to examine and compare soil moisture from satellite observations and Earth System Models. The methodology to derive this diagnostic is based on maximum likelihood estimator encoded into an iterative algorithm, which is applied to the soil moisture probability density function. This metric is applied to satellite data from the Advanced Microwave Scanning Radiometer for the Earth Observing System and global climate models data from the Coupled Model Intercomparison Project Phase 5 (CMIP5). Results show high soil moisture bimodality in transitional climate areas and high latitudes, potentially associated with land-atmosphere feedback processes. When comparing satellite versus climate models, a clear difference in their soil moisture bimodality is observed, with systematically higher values in the case of CMIP5 models. These differences appear related to areas where land-atmospheric feedback may be overestimated in current climate models.