Exposure to metal mixtures may affect children's health but the conclusions are controversial. We aimed to investigate the associations of metal mixture exposure with children's physical and behavioral development. 15 metals were detected in the urine samples of 278 preschoolers aged 3-6 years from eastern China. Multiple linear models and restricted cubic splines were used to evaluate dose-response relationships between single metal and children's physical and behavioral development. The Bayesian Kernel Machine Regression (BKMR) models, the weighted quantile sum (WQS) models and Quantile G-Computation were applied to evaluate the joint effects of metal mixtures. The results showed that arsenic (As) was negatively associated with z score of height for age (HAZ) in individual-metal models [β (95%CI): - 0.22 (-0.38, -0.06), P = 0.006]. Concerning children's behavioral development, multiple-metal models demonstrated a negative association with strontium (Sr) [β (95%CI): - 0.82 (-1.38, -0.26), P = 0.004], and a positive association with tin (Sn) [β (95%CI): 0.69 (0.16, 1.21), P = 0.010]. Notably, these associations remained significant or suggestive even after adjustments for multiple tests, sensitivity analyses, and application of different statistical models, including BKMR, WQS, and Quantile G-Computation. Furthermore, the study identified a negative joint effect of the metal mixture on HAZ, as demonstrated by BKMR and Quantile G-Computation models, with As playing an irreplaceable role in this observed impact. In summary, exposure to As appears to have adverse effects on HAZ, while exposure to Sn may hinder children's behavioral development. Conversely, exposure to Sr may have a protective effect on children's behavioral development. Additionally, the combined impact of metal mixtures is implicated in potentially impairing children's physical development, particularly in terms of HAZ.