The advancements in technologies such as location tracking, big data analytics, image processing, online retailing, and cloud computing, alongside innovations in artificial intelligence research, have propelled the adoption of machine learning (ML) models. These models surpass conventional econometric approaches in detecting patterns in complex, high-dimensional data, thereby enhancing predictive accuracy. In environmental economics, ML is increasingly utilized to analyze datasets from sensors, satellites, and texts, improving predictions, imputing missing values, uncovering counterfactual patterns for causal analysis, and gauging public sentiment via social media. We first present an overview of supervised, unsupervised, and causal ML models, discussing their applications so far in environmental economics, and evaluate their advantages and limitations. We then show that ML models have been used in four broad topics: (1) environmental policy evaluation, (2) environment and resource market analysis, (3) prediction of environmental outcomes, and (4) media analysis for environmental issues. We provide examples and report the gain from ML over conventional models to show the potentials of these methods in analyzing various topics. The review serves as a starting point for researchers seeking to explore the applications of ML in environmental economics.