Dysplasia affecting the oral mucosa has a malignant potential. The options for treatment of oral dysplasia are limited. Surgery remains the mainstay of management, with no strong evidence for medical treatments currently available. Histological grading alone does not accurately predict which dysplastic lesions will progress to cancer, which poses the clinician with difficult decisions regarding the most appropriate treatment and may lead to some patients being overtreated, with potentially unnecessary morbidity. The use of biomarkers may help to improve prediction of which cases are likely to transform. The efficacy of individual markers for prediction will be discussed in this paper. There is also a lack of consensus on the optimal frequency and duration of follow-up for patients with oral dysplasia.