The main objective of this theoretical review is to systematically analyze the type of International Classification of Sleep Disorders-3 (ICSD-3) diagnostic criteria by labeling each of them in order to propose an overview of the way in which the diagnostic criteria are organized. Labeling of diagnostic criteria using a rigorous iterative process of “aggregation” and “generalization” was conducted and inter-rater reliability calculation (Cohen's Kappa with three raters) was calculated. 241 criteria from 43 main sleep disorders of the ICSD-3 were labeled into nine types (Clinical manifestation 86.0% of sleep disorders, Objective markers 53.5%, Distress 30.2%, Disability 30.2%, Duration 30.2%, Frequency 58.1%, Age in 18.6%, Exclusion condition 81.4% and Associated condition 34.8%), with a high inter-rater reliability (Cohen's Kappa = 0.85). This analysis assumes that the structuring of the ICSD-3 diagnostic criteria is based on the Harmful Dysfunction Analysis (HDA). Some criteria correspond to the dysfunction part of the HDA while others refer to the harmful part. However, the approach does not seem to be homogeneous across the nosological classification. The use of a structured definition of sleep disorder and a framework to organize the ICSD diagnostic criteria is discussed with regard to the reliability and validity of criteria for diagnosing sleep disorders.