BACKGROUND: A questionable occlusal caries (QOC) lesion can be defined as an occlusal surface with no radiographic evidence of caries, but caries is suspected because of clinical appearance. In this study, the authors report the results of a 20-month follow-up of these lesions. METHODS: Fifty-three clinicians from The National Dental Practice-Based Research Network participated in this study, recording lesion characteristics at baseline and lesion status at 20 months. RESULTS: At baseline, 1,341 QOC lesions were examined; the treatment that was planned for 1,033 of those at baseline was monitoring (oral hygiene instruction, applying or prescribing fluoride or varnish, or both), and the remaining 308 received a sealant (n = 192) or invasive therapy (n = 116). At the 20-month visit, clinicians continued to monitor 927 (90 percent) of the 1,033 monitored lesions. Clinicians decided to seal 61 (6 percent) of the 1,033 lesions (mean follow-up, 19 months) and invasively treat 45 (4 percent) of them (mean follow-up, 15 months). Young patient age (< 18 years) (odds ratio = 3.4; 95 percent confidence interval, 1.7-6.8) and the lesion's being on a molar (odds ratio = 1.8; 95 percent confidence interval, 1.3-2.6) were associated with the clinician's deciding at some point after follow-up to seal the lesion or treat it invasively. CONCLUSIONS: Almost all (90 percent) QOC lesions for which the treatment planned at baseline was monitoring still were planned to undergo monitoring after 20 months. This finding suggests that noninvasive management is appropriate for these lesions. PRACTICAL IMPLICATIONS: Previous study results from baseline indicated a high prevalence of QOC lesions (34 percent). Clinicians should consider long-term monitoring when making treatment decisions about these lesions.