ObjectiveNew techniques have emerged to aid in preventing inadequate margins in oral squamous cell carcinoma (OSCC) surgery, but studies comparing different techniques are lacking. Here, we compared narrow band imaging (NBI) with fluorescence molecular imaging (FMI), to study which intraoperative technique best assesses the mucosal tumour margins. Materials and MethodsNBI was performed in vivo and borders were marked with three sutures. For FMI, patients received 75 mg of unlabelled cetuximab followed by 15 mg cetuximab-800CW intravenously-two days prior to surgery. The FMI borders were defined on the excised specimen. The NBI borders were correlated with the FMI outline and histopathology. ResultsSixteen patients were included, resulting in 31 NBI and 30 FMI measurements. The mucosal border was delineated within 1 mm of the tumour border in 4/31 (13 %) of NBI and in 16/30 (53 %) FMI cases (p = 0.0008), and within 5 mm in 23/31 (74 %) of NBI and in 29/30 (97 %) of FMI cases (p = 0.0048). The median distance between the tumour border and the imaging border was significantly greater for NBI (3.2 mm, range −6.1 to 12.8 mm) than for FMI (0.9 mm, range −3.0 to 7.4 mm; p = 0.028). Submucosal extension and previous irradiation reduced NBI accuracy. ConclusionEx vivo FMI performed more accurately than in vivo NBI in mucosal margin assessment, mainly because NBI cannot detect submucosal extension. NBI adequately identified the mucosal margin especially in early-stage and not previously irradiated tumours, and may therefore be preferable in these tumours for practical and cost-related reasons.