This article shows the long-term advantage of total resection of complex spinal cord lipomas over partial resection and over non-surgical treatment for children with asymptomatic lipomas. The classification, embryogenesis, and technique of total resection of complex lipomas are described. The 20-year outcome of 315 patients who had total resection is measured by overall progression-free survival (PFS, Kaplan-Meier), and by subgroup Cox multivariate hazard analysis for the influence of four variables: lipoma type, symptoms, prior surgery, and post-operative cord-sac ratio. These results are compared to 116 patients who underwent partial resection, and to two published series of asymptomatic lipomas followed without surgery. The PFS after total resection for all lipomas is 88.1% over 20 years vs. 34.6% for partial resection at 10.5 years (p < 0.0001). The PFS for total resection of asymptomatic virgin lipomas rose to 98.8% vs. 60% and 67% for non-surgical treatment. Partial resection also compares poorly to non-surgical treatment for asymptomatic lipomas. Multivariate analyses show that a low cord-sac ratio is the only independent variable that predicts good outcome. Pre-operative profiling shows the ideal patient for total resection is a young child with a virgin asymptomatic lipoma, who, with a PFS of 99.2%, is essentially cured. The technique of total resection can be learned by any neurosurgeon. Its long-term protection against symptomatic recurrence is better than partial resection and conservative management. The surgery should be done at diagnosis, except for asymptomatic small infants in whom surgery should be postponed till 6 months to minimize morbidity.