The aim of this study was to study the impact of a decision-making protocol for shunt type in the Norwood procedure for hypoplastic left heart syndrome. Our cohort extends from 2004 to 2016. In era 1 (pre-2008), there was no policy for the choice of Norwood shunt. In era 2 (post-2008), a standard protocol was implemented. The right ventricle (RV)-to-pulmonary artery conduit was utilized for low-birth weight patients (<2.5 kg). The right modified Blalock-Taussig Shunt (RBTS) was constructed for normal birth weight patients. The records of 133 consecutive operative patients with hypoplastic left heart syndrome anatomy between 2004 and 2016 were retrospectively reviewed. Survival risk factors were analysed using the Cox proportional hazards risk model. The Norwood procedure was performed at a mean age of 2.9 ± 1.9 days. Bidirectional cavopulmonary shunt was performed at a median age of 99 days (interquartile range 91-107). In era 1, 38.6% (22/57) of patients received the RBTS and 61.4% (35/57) of patients received the RV-to-pulmonary artery conduit. In era 2, 86.8% (66/76) of patients received the RBTS and 13.2% (10/76) of patients received the RV-to-pulmonary artery conduit. The actuarial survival to Fontan was 72.2% (96/133). Era 1 patients were more likely to die within the 1st year (hazard ratio = 2.310, P = 0.025). The shunt protocol may improve outcomes in high-risk patients, and we have demonstrated the reliability of the RBTS in low-risk patients. The short- and mid-term outcomes of our Norwood population justify the continued efforts to improve surgical and perioperative management.