With advances in Artificial Intelligence (AI) and the increasing volume of online educational data, Deep Learning techniques have played a critical role in predicting student’s academic performance. Recent developments have assisted instructors in determining the strengths and weaknesses of student’s academic achievement. This understanding will benefit from adopting the necessary interventions to assist students in improving their performance, helping at-risk students, and preventing dropout rates. In this review, 46 studies between 2019 and 2023 that apply one or more Deep Learning (DL) techniques, either alone or in combination with Machine Learning (ML) or Ensemble Learning techniques, have been reviewed and analyzed. Moreover, the review has examined utilizing datasets from public (MOOCs), private (LMSs), and other platforms. Four categories were used to group the features: demographic, previous and current academic performance, and learning behavior/activity features. Therefore, the analysis has demonstrated that the DNNs and CNN-LSTM models were the most commonly used techniques. Moreover, it has been noted that the studies that have used DL techniques such as CNNs, DNNs, and LSTMs, performed well by achieving high prediction accuracy above 90%; other studies achieved accuracy ranging between 60% and 90%. For datasets used within the reviewed studies, even though 44% of the studies have used LMSs datasets, it is found that OULAD was the most used dataset from MOOCs. For the grouped features, the results of the analysis indicate that the best features for prediction are the learning behavior and activity features, which outperform other feature categories.