Prospectively enrolled phenylketonuria patients (n=485) participated in an international Phase II clinical trial to identify the prevalence of a therapeutic response to daily doses of sapropterin dihydrochloride (sapropterin, KUVAN(®)). Responsive patients were then enrolled in two subsequent Phase III clinical trials to examine safety, ability to reduce blood Phenylalanine levels, dosage (5-20 mg/kg/day) and response, and bioavailability of sapropterin. We combined phenotypic findings in the Phase II and III clinical trials to classify study-related responsiveness associated with specific alleles and genotypes identified in the patients. We found that 17% of patients showed a response to sapropterin. The patients harbored 245 different genotypes derived from 122 different alleles, among which ten alleles were newly discovered. Only 16.3% of the genotypes clearly conferred a sapropterin-responsive phenotype. Among the different PAH alleles, only 5% conferred a responsive phenotype. The responsive alleles were largely but not solely missense mutations known to or likely to cause misfolding of the PAH subunit. However, the metabolic response was not robustly predictable from the PAH genotypes, based on the study design adopted for these clinical trials, and accordingly it seems prudent to test each person for this phenotype with a standardized protocol.