This cross-sectional study explores the relationships between children's F(-) exposure from drinking groundwater and urinary F(-) concentrations, combined with dental fluorosis (DF) in the Main Ethiopian Rift (MER) Valley. We examined the DF prevalence and severity among 491 children (10 to 15 years old) who are life-long residents of 33 rural communities in which groundwater concentrations of F(-) cover a wide range. A subset of 156 children was selected for urinary F(-) measurements. Our results showed that the mean F(-) concentrations in groundwater were 8.5 ± 4.1 mg/L (range: 1.1-18 mg/L), while those in urine were 12.1±7.3 mg/L (range: 1.1-39.8 mg/L). The prevalence of mild, moderate, and severe DF in children's teeth was 17%, 29%, and 45%, respectively, and the majority (90%; n=140) of the children had urinary F(-) concentrations above 3 mg/L. Below this level most of the teeth showed mild forms of DF. The exposure-response relationship between F(-) and DF was positive and non-linear, with DF severity tending to level off above a F(-) threshold of ~6 mg/L, most likely due to the fact that at ~6 mg/L the enamel is damaged as much as it can be clinically observed in most children. We also observed differential prevalence (and severity) of DF and urinary concentration, across children exposed to similar F(-) concentrations in water, which highlights the importance of individual-specific factors in addition to the F(-) levels in drinking water. Finally, we investigated urinary F(-) in children from communities where defluoridation remediation was taking place. The lower F(-) concentration measured in urine of this population demonstrates the capacity of the urinary F(-) method as an effective monitoring and evaluation tool for assessing the outcome of successful F(-) mitigation strategy in relatively short time (months) in areas affected with severe fluorosis.