Debate exists regarding the true pathogenicity of cerebral infundibula (CI). Pre-aneurysmal lesions and benign anatomical variants have both been proposed. In this study, we present the largest single cohort series on the natural history of CI. Retrospective review of prospective surveillance of 420 CI was undertaken in a single tertiary cerebrovascular centre. All CI diagnosed by a neuroradiologist, diagnosed on either a Magnetic resonance angiography (MRA), Computed tomography angiography (CTA) or Digital subtraction angiography (DSA) were eligible for inclusion. Imaging and demographic characteristics were recorded at baseline. CI growth and aneurysm transformation were the outcomes of interest. Groupwise comparison was conducted via Fischer exact testing. Kaplan Meir curves and Cox proportional hazard ratios were used to assess variables of interest with respect to time on surveillance. 402 patients with 420 CI were surveyed over 2418 infundibula-years. Eleven CI (2.62%) grew on surveillance, and three (0.7%) transformed into aneurysms. Median time to growth was 85 months (36-263) and median time to aneurysm transformation was 112 months (96-142). Of the CI that grew, male sex and CI >2 mm at diagnosis were significant predictors of growth (all p<0.05). Of the CI that grew in surveillance, 2/11 (18.2%) transformed into aneurysms (p=0.001). Aneurysm transformation occurred at a rate of 1.27 per 1000 infundibula years. CI growth on surveillance (p= 0.00016) and size at diagnosis (p=0.038) remained significant predictors of aneurysm transformation on Kaplan Meir curves. The transformation of a CI to an aneurysm occurs at a low rate. A history of growth on surveillance imaging represents significant risk for aneurysm transformation.