Our study was designed to prospectively determine whether or not unrelated cord blood transplantation (CBT) can produce outcomes comparable to related donor transplantation for children with β-thalassaemia. In 35 patients, 40 transplants were performed between October 2003 and September 2009. HLA matching at enrolment was 6/6 (n=8), 5/6 (n=16), 4/6 (n=27), or 3/6 (n=1) by low-resolution HLA-A, -B, and high-resolution DRB1. These patients received non-manipulated grafts without ex vivo expansion or T-cell depletion. The median number of nucleated and CD34+ cells infused was 7.8 × 10(7)/kg (range, 2.8-14.7 × 10(7)/kg) and 4.0 × 10(5)/kg (range, 1.7-19.9 × 10(5)/kg), respectively. The 5-year OS and thalassaemia-free survival after the first transplant were 88.3 and 73.9%, respectively. The cumulative incidence of TRM at 2 years was 11.7%. Fourteen patients developed chronic skin GVHD. Thirty patients were alive and transfusion-independent with a Lansky performance score ≥80% achieved between 6 and 76 months post transplant (median, 36 months). These data compare acceptably with the survival rates of related-donor BMT for thalassaemia and suggest that patients without an available HLA-compatible sibling but who have well-matched unrelated donors should also be considered for CBT.