From its characteristic clinical features, decrease of tyrosine hydroxylase (TH) in the terminal of the nigrostriatal (NS) dopamine (DA) neuron is considered the main lesion of HPD and the decrease of neopterin as well as biopterin in the cerebrospinal fluid suggested GTP cyclohydrolase I (GCH-I) as the responsible enzyme. By detecting the gene locus of GCH-I, Ichinose and his colleagues showed the abnormalities of GCH-I gene located on 14q 22.1 q22.2 as the cause of HPD. Since the first report of Ichinose et al, 11 mutations and frame shifts of the gene have been detected, in which the locus of abnormality differed among families but is identical in a family, but more than several families have been left with undetected abnormalities including those having linkage to 14q. However, the DNA of these families as well as those with detected gene abnormalities failed to synthesize GCH-I if inoculated with E. coli and the levels of GCH-I in mononuclear blood cells were below 20% of normal values in HPD patients while they were 37 and 38% in two asymptomatic carriers. Ratio of mutant mRNA of GCH-I gene was 28% in a patient and 8.3% in an asymptomatic case. These lines of evidence on GCH-I show HPD is a dominant inherited disorder with abnormalities of GCH-I gene. GCH-I is the limiting enzyme for synthesizing tetrahydrobiopterin (BH4), coenzyme transmitters for the synthesizing hydroxylases of aminergic neurotransmitters, but the affinity is the least for TH. This might cause a rather selective involvement of TH preserving serotonin synthesis un- or less affected. Fluoro-DOPA and [11C] racropride PET studies were normal in HPD. Studies of an autopsied case with dopa responsive dystonia, which was confirmed to have GCH-I gene abnormalities, neuropathologically revealed no abnormalities except for a decrease in melanin pigmentation in the substantia nigra and histochemically a decrease in TH enzyme activities and its protein only in the striatum. There was mild decrease of DA content, the interregional caudate/putamen and subregional rostrocaudal patterns which were similar to Parkinson disease, but subdivisionally different with predominant reduction in the ventral subdivision of the caudate nucleus. In the ventral part of the basal ganglia the striatal direct projection exists predominantly. Cases with recessive abnormalities of pteridin metabolism other than HPD, 6-pyruvoyl-tetra-hydropterin synthase (PSPS) deficiency and dihydropteridine reductase deficiency also show dystonia with diurnal fluctuation responding to levodopa, though not as marked as HPD. MPTP monkey studies revealed no involvement of striatal indirect pathway for peak dose dystonia. So it is suggested that in HPD, decrease of TH at the terminal of the NS-DA neuron due to partial reduction of GCH-I develops postural dystonia through the striatal direct projection in childhood with diurnal fluctuation depending on age and circadian variation of TH activities at the terminals.