The photosynthetic bacterium <
i>
Rhodobacter capsulatus<
/i>
normally photoproduces H<
sub>
2<
/sub>
as a by-product of its nitrogenase-catalyzed nitrogen-fixing activity. Such H<
sub>
2<
/sub>
production, however, is expensive from a metabolic perspective, requiring nearly four times as many photons as the equivalent algal hydrogenase-based system. Here we report the insertion of a Clostridium acetobutylicum [FeFe]-hydrogenase and its three attendant hydrogenase assembly proteins into an <
i>
R. capsulatus<
/i>
strain lacking its native uptake hydrogenase. Further, this strain is modified to fluoresce upon sensing H<
sub>
2<
/sub>
. The resulting strain photoproduces H<
sub>
2<
/sub>
and self-reports its own H<
sub>
2<
/sub>
production through fluorescence. Furthermore, this model system represents a unique method of developing hydrogenase-based H<
sub>
2<
/sub>
production in <
i>
R. capsulatus<
/i>
, may serve as a powerful system for in vivo directed evolution of hydrogenases and hydrogenase-associated genes, and provides a means of screening for increased metabolic production of H<
sub>
2<
/sub>
.