Measurements were made of net rates of CO2 assimilation in lichens at various ambient concentrations of CO2 in air and in helox (79% He, 21% O2). Because of the faster rate of CO2 diffusion in the pores of lichen thalli when filled with helox than when filled with air, a given net rate of assimilation was achieved at a lower ambient concentration of CO2 in helox. The differences were used to estimate resistances to diffusion through the gas-filled pore systems in lichens. The technique was first tested with five lichen species, and then applied in a detailed study with Ramalina maciformis, in which gas-phase resistances were determined in samples at four different states of hydration and with two irradiances. By assuming, on the basis of previous evidence, that the phycobiont in R. maciformis is fully turgid and photosynthetically competent at the smallest hydration imposed (equilibration with vapour at 97% relative humidity), and that, with this state of hydration, diffusion of CO2 to the phycobiont takes place through continuously gas-filled pores, it was possible also to determine both the dependence of net rate of assimilation in the phycobiont on local concentration of CO2 in the algal layer, and, with the wetter samples, the extents to which diffusion of CO2 to the phycobiont was impeded by water films. In equilibrium with air of 97% relative humidity, the thallus water content being 0.5 g per g dry weight, the resistance to CO2 diffusion through the thallus was about twice as large as the resistance to CO2 uptake within the phycobiont. Total resistance to diffusion increased rapidly with increase in hydration. At a water content of 2 g per g it was about 50 times as great as the resistance to uptake within the phycobiont and more than two-thirds of it was attributable to impedance of transfer by water. The influences of water content on rate of assimilation at various irradiances are discussed. The analysis shows that the local CO2 compensation concentration of the phycobiont in R. maciformis is close to zero, indicating that photorespiratory release of CO2 does not take place in the alga, Trebouxia sp., under the conditions of these experiments.