Background: It is unknown whether the neonatal tetrahydrobiopterin (BH4) loading test is adequate to diagnose long-term BH4 responsiveness in PKU. Therefore we compared the predictive value of the neonatal (test I) versus the 48-h BH4 loading test (test II) and long-term BH4 responsiveness. Methods: Data on test I (>1991, 20 mg/kg) at T = 8 (n = 85) and T = 24 (n = 5) were collected and compared with test II and long-term BH4 responsiveness at later age, with >= 30 % Phe decrease used as the cut-off. Results: The median (IQR) age at hospital diagnosis was 9 (7-11) days and the age at test II was 11.8 (6.6-13.7) years. The baseline Phe concentrations at test I were significantly higher compared to test II (1309 (834-1710) versus 514 (402-689) mu mol/L, respectively, P = 0.000). 15/85 patients had a positive test I T = 8. All, except one patient who was not tested for long-term BH4 responsiveness, showed long-term BH4 responsiveness. In 20/70 patients with a negative test I T = 8, long-term BH4 responsiveness was confirmed. Of 5 patients with a test I T = 24, 1/5 was positive at both tests and showed long-term BH4 responsiveness, 2/5 had negative results at both tests and 2/5 showed a negative test I T = 24, but a positive test II with 1/2 showing long-term BH4 responsiveness. Conclusions: Both a positive neonatal 8- and 24-h BH4 loading test are predictive for long-term BH4 responsiveness. However, a negative test does not rule out long-term BH4 responsiveness. Other alternatives to test for BH4 responsiveness at neonatal age should be investigated.