Alcohol withdrawal syndrome (AWS) complicated by seizures requires rapid symptom control while avoiding excessive sedation and other dose-related adverse drug reactions. This pilot randomized controlled trial evaluated whether pharmacogenetically guided optimization of pharmacotherapy could improve early clinical dynamics and tolerability in this high-risk inpatient population. Fifty inpatients with alcohol withdrawal state with convulsions (ICD-10 F10.31) were randomized 1:1 to pharmacogenetically guided treatment or control care. Pharmacogenetic testing was performed at admission in all patients; treating physicians in the intervention group received the true PGx report, whereas control physicians received a neutral report of identical format indicating wild-type genotypes and standard dosing. The primary endpoint was the area under the CIWA-Ar curve over 72 hours. Mean AUC CIWA-Ar was lower in the pharmacogenetic group than in controls (790 ± 175 vs 905 ± 185 score-hours; p = 0.032), with consistent non-parametric sensitivity analysis. Repeated-measures modelling showed a significant group-by-time interaction, with adjusted differences at 48 and 72 hours. The pharmacogenetic group also showed shorter time to stabilization, lower rescue medication burden, shorter hospitalization, lower UKU scores, and fewer clinically significant adverse events. These preliminary findings suggest that pharmacogenetically guided prescribing may improve tolerability and early withdrawal dynamics, but require confirmation in larger multicentre studies.