We aimed to examine the longitudinal associations of total cholesterol (TC), non-high-density lipoprotein cholesterol (non-HDL-C), high-density lipoprotein cholesterol (HDL-C), triglyceride, and low-density lipoprotein cholesterol (LDL-C) with carotid-femoral pulse wave velocity (cfPWV) and carotid intima-media thickness (cIMT) progression.We studied 1779, 15-year-old participants from the Avon Longitudinal Study of Parents and Children, UK birth cohort, followed up for 9 years. Fasting TC, non-HDL-C, HDL-C, triglyceride, and LDL-C were measured at 15, 17, and 24 years and age-categorized as normal, elevated, and dyslipidemia based on National Heart, Lung, and Blood Institute lipid guidelines. cfPWV and cIMT were measured at 17 and 24 years. Associations were examined using linear mixed-effect models. To simulate the treatment of dyslipidemia we conducted temporal inverse allocation model analyses.Among 1779 [49.9% female] participants, mean lipid levels and proportions at elevated or dyslipidemia categories increased from ages 15 through 24 years. Persistently elevated TC: effect estimate 0.026 mm; [95% CI 0.004 to 0.049; p = 0.024], elevated non-HDL-C, and elevated LDL-C were cumulatively associated with cIMT progression. Persistent borderline-low HDL-C: -0.027 mm; [-0.050 to -0.005; p = 0.019] and very-low HDL-C -0.035 mm; [-0.057 to -0.013; p = 0.002] levels were associated with cIMT progression. A temporal inverse allocation of elevated and dyslipidemic levels with normal lipid levels at age 17 years attenuated the associations of cumulative elevated TC, non-HDL-C, LDL-C, and low HDL-C with cIMT progression. Cumulative elevated lipids or dyslipidemia were not associated with cfPWV progression.Late adolescence is key to preventing, halting, and reversing dyslipidemic-related preclinical atherosclerosis progression, warranting universal lipid screening in the general pediatric population.