Microwave remote sensing emerged as a valuable technique for monitoring soil moisture (SM) due to the lower sensitivity to weather conditions and the ability to penetrate vegetation and surface layers. Given the latest advancements and changes in the field of microwave remote sensing for SM estimation, there is a need for a comprehensive and evidence-based evaluation of this technology from the extended published literature. Following the Center for Evidence-Based Conservation (CEBC) guidelines, we have selected and performed meta-analysis on 133 peer-reviewed research articles. The results show that microwave remote sensing has moderate to good accuracy in estimating SM (R2 ranged from 0.25 to 0.98, and RMSE from 0.005 to 0.141 m3/m3). Both active and passive sensors have unique spatial and temporal advantages, and while their use in combination is promising, the number of studies identified is limited, and the evidence for improved performance is inconclusive; therefore, further research and investigation are needed. Machine learning (ML) models, especially neural networks, significantly improved accuracy, especially when combined with semi-empirical models (median R2: 0.78 and median RMSE: 0.027 m3/m3). Regionally, microwave remote sensing has great potential for accurate monitoring in arid regions where SM is critical for resource management, with higher publication numbers and accuracy than in humid regions. As reported by the selected manuscripts, the performance inconsistency demonstrates the complexity of microwave remote sensing for SM estimation, which is mainly related to the sensor type, resolution level, and estimation method. This study delves into the state of the art of microwave remote sensing for SM measurement, highlighting performances, research gaps, and limitations.