Technologies for merging three-dimensional models of dental impressions include a computing device that generates multiple three-dimensional models that are each indicative of a dental impression of a user's dental arch. The models may be generated by scanning the dental impressions. The computing device determines whether a model is indicative of the complete anatomy of the user's arch and, if not, merges multiple models with a merge strategy to generate a merged model. The models may be merged by aligning geometry of the models, selecting geometry from one of the models using the merge strategy, and generating the merged model that includes the selected geometry. The merge strategy may include selecting from the model associated with the dental impression that includes the most detail of the user's anatomy or selecting from the model with the greatest depth.