This paper collected 134 groups of shallow groundwater samples from Luannan County from 2015 to 2023, constructed the groundwater quality evaluation model by the Random Forest method, and used the USEPA model to evaluate human health. The results showed that: (1) the groundwater quality in the study area was generally good, and only the samples of superwater accounted for 10.4%. The areas with poor water quality are concentrated in the southwest and north, and the central and eastern regions are good; (2) the importance evaluation of Random Forest shows that Mn2+ is the most important index affecting groundwater quality, mainly derived from mine drainage and unreasonable use of pesticides and fertilizers; (3) the proportion of children with HItotal greater than 1 reaches 64%, indicating that children are the most vulnerable population. Measures such as controlling pollution sources, regular monitoring, and health risk assessment are recommended to reduce risk.