The aim of the study was to investigate the developmental pattern of hypoplastic pulmonary artery (p.a.) bed augmented by systemic-to-pulmonary shunt in children with univentricular heart scheduled for Fontan surgery. For the study, a highly selected patient cohort was chosen (12 patients aged between 5 and 19 years; mean 9.5 years) with comparable initial morphological conditions of univentricular heart and hypoplastic p.a. bed, who after mandatory systemic-to-pulmonary shunt underwent Fontan procedure at time of normalization of pulmonary artery size. Further selection criteria were: normal pulmonary vascular resistance at time of Fontan procedure, competent a-v valve(s), and globally unimpaired ventricular function. All patients were grouped according to the preoperative pulmonary flow index (Qpi; L/min/m2 b.s.a.) measured immediately before Fontan operation: Group A: 1.5-2.5; B: 3.0-4.0; C: 4.0-5.0; D: > 6.0, and their cardio-pulmonary hemodynamic situation (Hb, SAsat%, Qp/Qs, PAP, Rp/Rs, EDVP, FS%, ventricular diastolic compliance (VC = EDVP/Qpi + Qsi) as well as the pulmonary artery size and area using standard (Nakata-index, McGoon-ratio) and a self designed computer assisted planimetric area calculation (PPAAI; cm2/m2 b.s.a.) analysed. Each patient underwent 1-3 shunt procedures, the mean shunt patency period for groups A, B, C and D was 12, 8.6, 5.3, and 4.5 years, respectively. The mean Nakata-index (283, 297, 324, 405 in groups A-D) and the McGoon-ratio (2.0, 2.2, 2.8, 3.3 in groups A-D) correlated with the Qp index, reflecting flow dependent development of pulmonary artery bed. No correlation was found between Qpi and PPAAI (47, 40, 41 and 47 in group A-D). The VC/Qp relation showed an inversely proportional pattern with values 2.3, 1.0, 0.8, 0.7 for corresponding groups A-D, the lowest VC in group A correlated with polyglobulic status (Hb- values; g/dl): 21.3 in A vs 19.8, 18.0 and 16.5 in B-D) and mean arterial SAsat-values (77% in A vs 83%, 84% and 89% in B-D). In conclusion, in our highly selected patient cohort, the development of p.a. size was strongly flow-dependent, and patients with restrictive pulmonary flow needed an approximately threefold longer time period to normalize their p.a. size compared to those with excessive flow. In patients with restrictive pulmonary flow, the Nakata-index underestimated the degree of development of the pulmonary artery system, probably due to the distortion of the proximal p.a. segment. In consequence, in these patients the normalization of the p.a. bed and thus suitability for the Fontan procedure probably occurred much earlier. Based on our observations and those of others, in patients with excessive flow the normalization of p.a. bed, provided it occurs within 3-4 years, seems not necessarily to be associated with a deterioration of ventricular function.