A thermosiphon photobioreactor (TPBR) can potentially be used for biohydrogen production, circumventing the requirement for external mixing energy inputs. In this study, a TPBR is evaluated for photofermentative hydrogen production by Rhodopseudomonas palustris (R. palustris). Experiments were conducted in a TPBR, and response surface methodology (RSM), varying biomass concentration, and light intensity and temperature were employed to determine the operating conditions for the enhancement of both hydrogen production as well as biomass suspension. Biomass concentration was found to have had the most pronounced effect on both hydrogen production as well as biomass suspension. RSM models predicted maximum specific hydrogen production rates of 0.17 mol m−3h−1 and 0.21 mmol gCDW−1h−1 at R. palustris concentrations of 1.21 and 0.4 g L−1, respectively. The experimentally measured hydrogen yield was in the range of 45 to 77% (±3.8%), and the glycerol consumption was 8 to 19% (±0.48). At a biomass concentration of 0.40 g L−1, the highest percentage of biomass (72.3%), was predicted to remain in suspension in the TPBR. Collectively, the proposed novel photobioreactor was shown to produce hydrogen as well as passively circulate biomass.