Study objective: Although adjunctive intravenous bicarbonate therapy is commonly recommended for children with severe diabetic ketoacidosis (DKA), no studies assessing clinical outcome with this therapy have ever been performed. Our objective was to determine whether bicarbonate therapy influenced outcome for pediatric DKA. Methods: The study was a retrospective consecutive case series of 147 admissions for severe DKA (initial pH ≤7.15 and glucose concentration ≤300 mg/dL [16.7 mmol/L]) in 106 children during a 16-year period at a tertiary University medical center. Descriptive statistics were applied to the 147 admissions. The first patient admitted with DKA was then selected for each of the 106 children, and clinical and laboratory data were compared between subjects who did and did not receive bicarbonate. Multivariate and matched pair analyses were performed to control potentially confounding variables. Results: Fifty-seven of the 147 patients admitted with DKA (39%) were successfully treated without bicarbonate, including 9 with a pH of 7.00 or less and one with a pH of 6.73. The frequency of complications was comparable between bicarbonate and nonbicarbonate groups (4% versus 2%, P=1.00). The mean duration of hospitalization for children receiving bicarbonate was 23% (16 hours) longer than children who did not receive bicarbonate in the multivariate analysis (P=.07) and 37% (22 hours) longer in the matched pair analysis (P=.01). The mean rate of metabolic recovery by three distinct measures was similar between groups, and the sample had 80% power to detect differences of 14% to 29% in these measures. Conclusion: We found no evidence that adjunctive bicarbonate improved clinical outcome in children with severe DKA. The rate of metabolic recovery and complications were similar in patients treated with and without bicarbonate, and prolonged hospitalizations were noted in the bicarbonate group. We conclude that adjunctive bicarbonate is unnecessary and potentially disadvantageous in severe pediatric DKA.