Tryptophan hydroxylase (TPH) carries out the 5-hydroxylation of L-Trp, which is the rate-limiting step in the synthesis of serotonin. We have prepared and characterized a stable N-terminally truncated form of human TPH that includes the catalytic domain (Delta90TPH). We have also determined the conformation and distances to the catalytic non-heme iron of both L-Trp and the tetrahydrobiopterin cofactor analogue L-erythro-7,8-dihydrobiopterin (BH2) bound to Delta90TPH by using 1H NMR spectroscopy. The bound conformers of the substrate and the pterin were then docked into the modeled three-dimensional structure of TPH. The resulting ternary TPH-BH2-L-Trp structure is very similar to that previously determined by the same methods for the complex of phenylalanine hydroxylase (PAH) with BH2 and L-Phe [Teigen, K., et al. (1999) J. Mol. Biol. 294, 807-823]. In the model, L-Trp binds to the enzyme through interactions with Arg257, Ser336, His272, Phe318, and Phe313, and the ring of BH2 interacts mainly with Phe241 and Glu273. The distances between the hydroxylation sites at C5 in L-Trp and C4a in the pterin, i.e., 6.1 +/- 0.4 A, and from each of these sites to the iron, i.e., 4.1 +/- 0.3 and 4.4 +/- 0.3 A, respectively, are also in agreement with the formation of a transient iron-4a-peroxytetrahydropterin in the reaction, as proposed for the other hydroxylases. The different conformation of the dihydroxypropyl chain of BH2 in PAH and TPH seems to be related to the presence of nonconserved residues, i.e., Tyr235 and Pro238 in TPH, at the cofactor binding site. Moreover, Phe313, which seems to interact with the substrate through ring stacking, corresponds to a Trp residue in both tyrosine hydroxylase and PAH (Trp326) and appears to be an important residue for influencing the substrate specificity in this family of enzymes. We show that the W326F mutation in PAH increases the relative preference for L-Trp as the substrate, while the F313W mutation in TPH increases the preference for L-Phe, possibly by a conserved active site volume effect.