Crop growth conditions are key information sources for estimating and forecasting crop yields, which are of great value to food security and trade. With the continuous appearance of high spatial and temporal resolution remote sensing data, the remote sensing data have presented obvious characteristics of big data. Therefore, crop growth monitoring and yield estimation based on deep learning has become one of the important means to guide agricultural production. The research status of deep learning at the regional scale was investigated, which focused on the development of model samples and model structure. Among them, the model samples were summarized through two aspects of sample construction and sample augmentation. The progress of the deep learning model structure of convolutional neural network (CNN), recurrent neural network (RNN), and their optimized structures and model interpretability were also summarized. Besides, the latest progress of crop growth monitoring and yield estimation at field scale at home and abroad was elaborated from two aspects: unmanned aerial vehicle (UAV) platform and satellite platform. Finally, the existing problems and the future perspective were analyzed and discussed, including improving the limitation of small samples through region-based and parameter-based transfer learning, the organic combination of deep learning model and crop growth model to improve the interpretability of the model, and the combination of UAV platform and satellite platform to ensure the precision of scale conversion in the process of spatio-temporal fusion, which can further explore the potential of deep learning in crop growth monitoring.