Tooth agenesis is the most common developmental dental anomaly. Absence of one or two permanent teeth is found in the majority of affected subjects. Very few patients suffer severe tooth agenesis. Recent studies revealed that WNT10A gene mutations caused syndromic and isolated severe tooth agenesis. In this study, to determine the contribution of WNT10A variants in different severities of tooth agenesis, we investigated the association between WNT10A variants and non-syndromic tooth agenesis in a Chinese population consisting of 505 tooth agenesis patients and 451 normal controls. Twenty-three novel non-synonymous variants were identified. WNT10A variants were detected in 15.8 % (75/474) of patients with 1-3 missing teeth and 51.6 % (16/31) of patients with 4 or more missing teeth. As compared with a frequency of 3.1 % in individuals with full dentition, variant allele frequencies were significantly elevated in both groups with tooth agenesis (p values of 1.00 × 10(-6) and 3.89 × 10(-23), respectively). Our findings showed that WNT10A variants were associated with non-syndromic tooth agenesis from mild to severe tooth agenesis, and the more severe tooth agenesis, the stronger association. Biallelic genotypes of WNT10A variants may have a pathogenic effect on tooth development. Presence of a single variant allele would be predisposing for causation with low penetrance. Together with WNT10A variant, there should be other genetic or environmental factors leading to biallelic variant-related variable clinical manifestations and single allele variant-related low penetrance. The frequent missing tooth positions in the WNT10A-related cases were consistent with that in the general population, suggesting WNT10A plays a critically important role in the etiology of general tooth agenesis.