Purpose: The purpose of this single-blinded, 2-centre, randomized controlled trial was to test if near-infrared (NIR) autofluorescence image guidance for parathyroid gland (PG) detection during total thyroidectomy can reduce the incidence of hypoparathyroidism in both malignant and benign cases. Method: Patients admitted for primary or completion total thyroidectomy were randomized to either the NIR intervention group or the standard care NONIR (no near infrared) group. The primary endpoint was the rate of hypoparathyroidism at the 3-month follow-up, defined as hypocalcemia and inappropriately low parathyroid hormone levels and/or continuous treatment with active vitamin D. The secondary endpoint was the PG identification rate. Results: A total of 147 patients were included of whom 73 were allocated to NIR. Primary or completion thyroidectomy was conducted in 84 and 63 cases, respectively. A total of 130 completed 3 months follow-up. Postoperative hypoparathyroidism in the NIR group at 12 h, 1 month and 3 months was, respectively, 31.8, 14.1, 6.5% compared with 35.9, 18.9, 11.8% in the NONIR group (all p > 0.46). In the NIR group, the identification rate of PGs was 69.5% (146 of 210 PGs), and 9% (19 of 210 PGs) were identified only due to additional use of NIR. For 15 out of 69 patients (21.7%) additionally PGs was found. Conclusion: Hypoparathyroidism was nominally less frequent in the NIR group, although not statistically significant. Further studies are needed to confirm if NIR may be a supportive PG identification tool to minimize the number of PG which would have been otherwise missed, especially during more complicated thyroid procedures. Trial registry: ClinicalTrials.gov: NCT04193332. Registration date: 16.08.2019.