Introduction: Prediction of unerupted permanent teeth is an essential part of orthodontic diagnosis and treatment planning. This prediction is done by mixed dentition space analysis based on the estimation of mesiodistal dimensions of unerupted permanent canine and premolars from already erupted permanent teeth. Permanent mandibular incisors are most commonly used for prediction. Recent literature reveals that mandibular incisors are not accurate predictors and other independent variables have been introduced to make a more accurate and precise prediction. The objective of this paper was to evaluate the literature in light of a variety of independent variables and their predictive accuracy. Methods: Electronic databases such as MEDLINE, PubMed, Scopus, Embase, Web of Science, CINAHL Plus were searched to identify articles published until September 2021. Results: The search resulted in a total of 1098 articles, of which 24 papers met our inclusion criteria and were included in this review. Articles using permanent mandibular incisors only as a predictor were excluded during the eligibility assessment. The results show that various independent variables including mesiodistal and vestibulo-oral dimensions of permanent maxillary and mandibular incisors and molars, molar basal arch length, intermolar distance, maxillary and mandibular arch and gender have been used as predictors to more accurately determine mesiodistal width of unerupted canine and premolars in different populations. Conclusion: Ethnic tooth size variations strongly emphasize the need to determine which independent variable gives a more accurate prediction of unerupted permanent teeth to develop a population-specific prediction model. This will play a significant role in managing space problems and developing malocclusions.