OBJECTIVE Clinicians have difficulty assessing and monitoring early occlusal caries. Traditional clinical exam and radiographs are unable to detect the subtle alterations in enamel indicative of de- or re-mineralization, particularly under dental sealants. Although clinicians have used laser fluorescence (LF) to address this gap, this modality has demonstrated weak correlation with histology. The International Caries Detection and Assessment System (ICDAS-II) has demonstrated high sensitivity and specificity for caries detection, but since it is based on visual assessment, it is of no use in areas beneath the most commonly used dental sealants which are opaque. Optical coherence tomography (OCT) is an emergent assessment tool which has demonstrated great promise in detecting and quantifying caries, including areas beneath commonly used dental sealants and composites. However, OCT has not yet been widely integrated into clinical dental practice, perhaps because OCT imaging does not provide an easily accessible diagnostic outcome for clinicians. The objective of this ex vivo study was to use OCT-images of sound and carious occlusal surfaces in combination with a simple algorithm to compare the caries detection ability of OCT with tools clinicians may be more familiar with (LF and radiography), and with an established valid and reliable clinical assessment tool (ICDAS-II). METHODS One hundred twenty extracted teeth with sound or naturally carious occlusal surfaces were imaged with OCT, LF, radiography, and examined clinically with the ICDAS-II. Teeth were randomized to one of two dental sealants recommended for use with LF. A novel simple algorithm was used to interpret OCT-based images. The accuracy of caries severity assessments of the OCT-based diagnosis, LF, ICDAS-II, and digital radiography were compared to the 4-point histological analysis gold standard. RESULTS OCT and ICDAS-II caries severity assessments demonstrated high sensitivity (94.0%; 92.3%) and specificity (85.0%; 83.3%), LF demonstrated low sensitivity (65.2%) but high specificity (97.6%), and digital radiography demonstrated low sensitivity (67.1%) with moderate specificity (79.5%) on unsealed occlusal surfaces. OCT-based caries severity assessments of sealed teeth demonstrated high specificity (97.6%), sensitivity (89.9%), excellent positive predictive value (98.6%), and negative predictive value (83.3%). Despite our use of LF recommended dental sealants, in the presence of sealants, LF assessment of caries severity demonstrated high sensitivity (95.1%), but extremely low specificity (10.3%), positive predictive value (68.8%), and negative predictive value (50.0%). CONCLUSIONS This study found that OCT-based imaging combined with a simple diagnostic algorithm accurately assessed the severity of natural early caries on occlusal surfaces in extracted teeth both in the absence and presence of dental sealant. The findings of this study support the clinical use of OCT imaging for assessment and monitoring progression of early non-cavitated caries lesions on occlusal surfaces including areas under dental sealants.