In this study the photobiocatalytic hydrogen production system was photoelectrochemically examined. Preliminary experiments with a mixed slurry system revealed the following five facts: direct, inter-phase, electron transfer from photocatalyst to enzyme in the absence of electron relay was a rate-determining step; the enzyme was deactivated by irradiated light (half after 10 days); physical properties (crystallinity, specific surface area, pore volume and radius) of the photocatalysts, obtained by two preparation methods (sol-gel and hydrothermal), were seldom correlated with the production rate; chemical properties were postulated to have been affected; and each material needed different reaction conditions such as pH and temperature. For the later salt-bridged system, the I-V relation for each prepared photocatalyst was measured (Vfb in the range of ca. .0.9~.1.0 and Voc close to .1.1 V vs. Ag/AgCl in saturated KCl). The change in the amount of evolved hydrogen was also independently checked according to external bias (critical bias around 1.0 V). Then both reaction components were connected through a salt bridge to initiate light-induced hydrogen production for the hybrid system. The H2 production exhibited an exponential trend and the Tris-HCl buffer showed the highest rate. A feasible reaction pathway was proposed.