Purpose <p>This study aimed (1) to investigate the spatiotemporal distribution of rare earth elements (REEs) in the surface sediments of Xinglin Bay watershed; (2) to evaluate the pollution levels and ecological risks posed by REEs in the sediments; and (3) to quantitatively identify the sources of REEs in the surface sediments by combining Sr-Nd isotopes and Bayesian mixing models.</p> Methods <p>This study measured the concentrations of REEs in sediments during normal season and flood season in the Xiamen Xinglin Bay area of China, and calculated the geo-accumulation index, pollution load index, and potential ecological risk index. In addition, based on the characteristic parameters of REEs, a combination of isotope tracing and Bayesian mixing models were applied to analyze the sources of REEs.</p> Results <p>The concentrations of all REEs in the sediments were higher than the background values of local soil during both seasons, and the concentrations in normal season were higher than those in flood season. The geo-accumulation index indicated no obvious REEs pollution in the surface sediments. The pollution load index indicated slight REEs pollution in the surface sediments. However, the potential ecological risk indexes of the REEs were moderate, and the sum contribution rates of Lu, Tm, and Tb were more than 42%, indicating the ecological risk of these elements cannot be ignored. The Sr-Nd isotopes combined with Bayesian mixing models demonstrated that the REEs in the sediments mainly came from natural sources (47.78%, 29.84%), agricultural activities (39.83%, 56.60%), and industrial activities (12.39%, 13.57%) during normal and flood seasons respectively.</p> Conclusion <p>The study area has a moderate potential ecological risk from the presence of REEs in surface sediments. The REEs in the surface sediments were mainly from natural sources, agricultural activities and industrial activities.</p> Graphical abstract <p></p>