Wildfire is a growing concern worldwide with significant impacts on human lives and the environment. This study aimed to provide an overview of the current trends and research gaps in wildfire prediction by conducting a bibliometric analysis of papers in the Web of Science and Scopus databases. CiteSpace was employed to analyze the co-occurrence of keywords, identify clusters, and detect emerging trends. The results showed that the most frequently occurring keywords were “wildfire”, “prediction”, and “model” and the top three clusters were related to “air quality”, “history”, and “validation”. The analysis of emerging trends revealed a focus on vegetation, precipitation, land use, trends, and the random forest algorithm. The study contributes to a better understanding of the research trends and gaps in wildfire prediction and provides recommendations for future research, such as incorporating new data sources and using advanced techniques.