The authors describe an experimental system and techniques for use in simultaneous pollutant exposure experiments and gaseous flux determinations. The system uses flexible Teflon bag-like chambers to enclose entire individual branches of young trees. Five gaseous fluxes (CO2, H2O, SO2, O3, and H2O2) are measured once per hour for each of two branches. Techniques for determining chamber surface loss corrections and methods of estimating needle surface area of conifers are described. A three-week continuous test was run in order to evaluate system performance. Inlet concentrations of the five gases were extremely stable, with uncertainties of <2.2% in all cases and day-to-night differences of <4%. Comparison of two branch chambers, one with clean air and one with the added pollutant gases H2O2, SO2, and O3, showed that between-chamber differences were <2% for the average inlet concentrations of H2O and CO2. Pollutant losses to chamber surfaces were significant but were generally smaller than losses to the branch itself. Chamber loss corrections were different for daytime and nighttime conditions and also evolved as the three-week experiment proceeded.