Proportional-integral-plus (PIP) control is employed to maintain gas concentration in a small-scale free air carbon dioxide enrichment (FACE) system. FACE systems are designed to produce controlled concentrations of elevated carbon dioxide, or other atmospheric gases, enabling plant growth experiments to be carried out for in situ vegetation without the use of chambers or other enclosures. Current FACE systems employ control algorithms based on classically derived two- or three-term control laws with manually tuned parameters. However, small FACE plots are more susceptible to turbulent eddies than larger scale systems, making control of concentration particularly difficult. The research described in the present paper employs data from planned FACE experiments to develop PIP control algorithms exploiting model-based predictive control action. Initial trials utilizing this approach yield good results for a small-scale FACE system operating in an uncut arable meadow.