Aim: To quantify and compare the variation in the dimensions of the atlas vertebra among patients with classes I, II, and III skeletal bases on a lateral cephalogram. Materials and methods: Lateral cephalograms of 75 patients aged 19–43 years were evaluated. Subjects were divided into three groups (25 each) based on their A point, nasion, B point (ANB) angle. Group I included class I subjects with ANB angle 0–4^°, group II included class II subjects with ANB angle >4^°, and group III included class III subjects with ANB of <0^°. Three linear measurements, that is, the anteroposterior dimension, height of the slimmest part of the posterior neural arch, and height of the dorsal arch were used to assess the dimensions of the atlas vertebra. The data were recorded, tabulated, and statistically analyzed. Results: Mean difference in the anteroposterior dimension of the atlas vertebra was statistically significant with the highest dimension in the class II group followed by classes I and III (p < 0.001) whereas, the mean height of the dorsal arch was found to be the highest in the class III group followed by classes I and II (p < 0.000). Conclusion: There are variations in the anteroposterior dimension and height of the dorsal arch of the atlas vertebra among patients with classes I, II, and III skeletal bases. Clinical significance: Dimensional measurements of the atlas vertebra and its specific variations in patients with different skeletal base patterns performed on a lateral cephalogram itself give an endowment to study the morphology of the atlas vertebra in routine clinical status quo without the requisite of any added diagnostic investigations such as cone-beam computed tomography (CBCT). Understanding of such variances can be integrated during diagnosis and treatment planning to take suitable preventive and interceptive procedures in any developing patterns of malocclusion.