Decision making for infants presenting with HLHS is complex and inhibited by an absence of comparative studies. Transplantation has a high success rate and good functional long term outcome, carries the unknown risks of long term immunosuppression and is limited by donor availability. The “Norwood” : procedure has an improving survival rate in experienced hands, but also carries unknown long term risks particularly neurodevelopmental outcome. Decision analysis has been developed as a way of making complex decisions using the best available information and should be ideally suited to this situation. Methods: A literature search for the latest information on survival rates was performed in September 1997. The initial values entered into the model were: survival of stage 1 Norwood, 85%; survival of stage 2, 95%; and of stage 3, 90%. Transplant availability 30% in the first month of life and 40% of the remaining infants in the next 5 months; immediate survival, 85% and 5 y survival, 90% of immediate survivors. Variations in the values assigned to the parameters were examined to determine the effects of improving donor availability or improving surgical results. A combined approach was also evaluated in which infants would be listed for transplant at birth and undergo a stage 1 Norwood procedure at 1 month of age if they have not received a transplant by this time; infants would then remain on the transplant list until 6 months at which time a stage 2 procedure is performed and the infant taken off the transplant list. Results: With the staged surgical repair approach alone the model gave a 5 y survival rate of 52% similar to the best recently reported results of 58% (Bove and Lloyd 1996). For infants placed on a transplant list alone, the model calculates 26% 5 y survival. The combined approach using the same parameters gives an expected 5 y survival of 61%. If donor availability improves, so that 80% of infants could be transplanted before 6 months, the 5 y survival of transplantation improves to 56% and survival rate of the combined approach improves to 67%. Improvements in Norwood results to a feasible maximum 5 y survival of 66% also improved the outcome of infants assigned to the combination protocol to 71% 5 y survival. Discussion. The combined approach performed best in this analysis, a result which was quite robust to changes in the assumptions underlying the model. Such an approach is currently offered in few centers. Long term neurodevelopmental follow up is essential for further improvements in decision making.