Floods are among the most destructive natural disasters, causing significant economic and human losses in Iran annually. Accurate modeling of this phenomenon has always faced challenges due to the nonlinear and complex nature of the rainfall-runoff process. In this review study, nearly 70 reputable scientific papers, including approximately 38 studies with individual models and 32 studies with hybrid models conducted between 2003 and 2025 in various watersheds across Iran, have been reviewed and analyzed. The main objective is to compare the performance of individual and hybrid models in flood and runoff simulation and to identify successful approaches in Iran. The findings indicate that individual models such as HEC-HMS, SWAT, SWMM, ANN, ANFIS, and WRF, despite their simplicity, face serious weaknesses in estimating peak discharges and extreme flood events, while hybrid models have performed better in over 90% of cases, with average accuracy improvement ranging from 15 to 30%. Among the hybrid approaches, three main categories including wavelet-based models (Wavelet-ANFIS, WSVR), metaheuristic optimization models (ANFIS-ICA, SVR-HHO), and physical-machine learning hybrid models (HEC-HMS+LSTM, WRF+HEC-HMS+HEC-RAS) have shown the most successful results in mountainous and semi-arid watersheds of Iran. Models such as Random Subspace J48, SA-ABM, and Deep Q-Learning have also shown significant performance in flood hazard zoning. Based on the results, hybrid models with 15-30% accuracy improvement have taken an effective step toward increasing the reliability of flood prediction in Iran. However, challenges such as lack of long-term data, uncertainty analysis, computational complexity, lack of standardization in model evaluation, and limited studies in southern and eastern watersheds remain as the main obstacles. Targeted investment in data infrastructure, specialized training, establishment of a national flood forecasting system based on hybrid models, and attention to uncertainty analysis are the main recommendations of this study.