Sample size calculation is a fundamental step for the validity and the usefulness of results from a study. Nevertheless, as demonstrated by some papers in the literature, such calculation is often ignored. Despite the lack of papers on this topic, it is probable that this shortcoming also affects studies on temporomandibular disorders. Therefore, the aim of this paper was to provide some basic rules to calculate the sample size necessary for different types of studies, both longitudinal and transversal, on those pathologies. Some examples of the application of such rules for different types of studies have also been provided, in order to make a full comprehension easier. In fact, the systematic application of those rules is strongly requested for the effective usefulness of results. Furthermore, an analysis of the statistical power of past studies on temporomandibular disorders could be useful to evaluate if our epidemiological and clinical-therapeutic knowledge of temporomandibular disorders is effectively based upon studies conducted with the appropriate sample size.