The determination of clinico-surgical predictors of carotid body tumour (CBT) metastasizing and/or recurrence. Patients with CBT but without a history of its previous treatment who had received inpatient medical aid at a tertiary surgical centre during the period 2011-2021 were retrospectively reviewed to obtain baseline data. Primary endpoints of the study were CBT metastasizing and/or local recurrence after radical surgery (composite endpoint) and overall survival. Secondary endpoint of the study was primary patency of carotid arteries that had been repaired/reconstructed as a result of CBT removal. Suspected predictors of CBT metastasizing/recurrence were young age of CBT onset, positive family history, multifocal paraganglioma, Shamblin type III CBT, carotid bifurcation and/or internal carotid artery (ICA) reconstruction as a result of CBT excision. To detect CBT metastasizing/recurrence, prospective imaging-based whole-body screening was suggested to patients. Fifty-three patients met the eligibility criteria. The overall survival rate was 100% at a median follow-up time of 68 months (interquartile range [IQR] 27‒104.5 months). The primary patency of repaired carotid arteries (n = 9) was 88.9% at a median follow-up time of 99 months (IQR 15.5‒116.5 months). Thirty-nine patients underwent the prospective whole-body screening. Eight (20.5%) patients were diagnosed with metastatic disease (n = 6) or local recurrence (n = 2). Patients with CBT metastasizing/recurrence were younger at age of CBT onset (p = .011), more frequently had Shamblin type III CBT (p = .004) and the necessity for carotid artery repair/reconstruction (p = .041) than patients without metastasizing/recurrence. These two groups were similar in terms of gender (p = .682), the prevalence of multifocal paraganglioma (p = .167) and positive family history (p = .180). Only the belonging of a CBT to Shamblin type III was defined as a statistically significant predictor of CBT metastasizing/recurrence in both univariable and multivariable binary logistic regression analyses (odds ratio 30; 95% confidence interval: 2.649-339.739; p = .006 and odds ratio 75.409; 95% confidence interval: 1.082-5254.557; p = .046, respectively). Multifocal paraganglioma as a predictor exhibited a strong tendency towards statistical significance (p = .09 and p = .072, respectively). Other suspected predictors didn't reach the level of statistical significance either in univariable (p = .151 for positive family history) or in multivariable (p = .211 for age of CBT onset; p = .734 for carotid artery reconstruction/repair) binary logistic regression analyses. Shamblin type III CBT has high potential for metastasizing and recurrence. Patients with Shamblin type III CBT and/or multifocal paraganglioma should be considered for genetic testing as well as preoperative whole-body radiological imaging and long-term postoperative whole-body imaging-based follow-up.