Abstract Aims To investigate relationships between takeaway food and sugar-sweetened beverage (SSB) consumption with cardiometabolic phenotypes during childhood and mid-adulthood. Method Design: Cross-sectional Child Health CheckPoint within the national population-representative Longitudinal Study of Australian Children. Participants: 1838 children (mean age 11.5 years; 49.1% female) and 1846 adults (mean age 43.7 years; 87.6% female). Exposures: Self-reported takeaway food and SSB consumption (‘frequent’: ≥ weekly). Outcomes: Functional (pulse wave velocity (PWV), blood pressure (BP)) and structural (carotid intima-media thickness, retinal microvascular calibre) preclinical cardiovascular phenotypes; lipids (total cholesterol, low-density lipoprotein (LDL), high-density lipoprotein (HDL), triglycerides). Analysis: Linear regression (exposure: takeaway or SSB consumption, individually or together) adjusted for age, sex and socio-economic position; and mediation analysis for body mass index (BMI). Results Associations were small among children (standardized mean difference (SMD) ≤0.15). In adults, associations were stronger with functional, but not structural, cardiovascular phenotypes and lipids, particularly for frequent takeaway food consumption (e.g. PWV (0.20 m/s; 95% confidence interval (CI) 0.03 to 0.37); systolic (3.3 mmHg; 95% CI 1.3 to 5.3) and diastolic BP (1.4 mmHg; 95% CI 0.2 to 2.6); LDL (0.10 mmol/L; 95% CI 0.02 to 0.18); HDL (−0.14 mmol/L; 95% CI −0.19 to −0.10) and triglycerides (0.30 mmol/L; 95% CI 0.12 to 0.48)]. BMI mediated associations between takeaway food consumption and PWV, BP, HDL and TG (proportion of mediation 34% to 75%), while mediation effects were smaller for SSB consumption. Conclusions Frequent takeaway food consumption in adults was associated with adverse blood lipids and vascular function (mainly via BMI). Lack of strong associations in children highlights opportunities for prevention.