Objective—To compare efficacy and safety of treatment with phenobarbital or bromide as the first-choice antiepileptic drug (AED) in dogs. Design—Double-blinded, randomized, parallel, clinical trial. Animals—46 AED-naïve dogs with naturally occurring epilepsy. Procedures—Study inclusion was based on age, history, findings on physical and neurologic examinations, and clinicopathologic test results. For either phenobarbital treatment (21 dogs) or bromide treatment (25), a 7-day loading dose period was initiated along with a maintenance dose, which was adjusted on the basis of monthly monitoring. Efficacy and safety outcomes were compared between times (baseline and study end [generally 6 months]) and between drugs. Results—Phenobarbital treatment resulted in eradication of seizures (17/20 [85%]) significantly more often than did bromide (12/23 [52%]); phenobarbital treatment also resulted in a greater percentage decrease in seizure duration (88 ± 34%), compared with bromide (49 ± 75%). Seizure activity worsened in 3 bromide-treated dogs only. In dogs with seizure eradication, mean ± SD serum phenobarbital concentration was 25 ± 6 μg/mL (phenobarbital dosage, 4.1 ± 1.1 mg/kg [1.9 ± 0.5 mg/lb], PO, q 12 h) and mean serum bromide concentration was 1.8 ± 0.6 mg/mL (bromide dosage, 31 ± 11 mg/kg [14 ± 5 mg/lb], PO, q 12 h). Ataxia, lethargy, and polydipsia were greater at 1 month for phenobarbital-treated dogs; vomiting was greater for bromide-treated dogs at 1 month and study end. Conclusions and Clinical Relevance—Both phenobarbital and bromide were reasonable first-choice AEDs for dogs, but phenobarbital was more effective and better tolerated during the first 6 months of treatment.