Responses of leaf stomatal conductance to light, humidity and temperature were characterized for winter wheat and barely grown at ambient (about 350 μmol mol−1 in the daytime), ambient + 175 and ambient + 350 μmol mol−1 concentrations of carbon dioxide in open-topped chambers in field plots over a three year period. Stomatal responses to environment were determined by direct manipulation of single environmental factors, and those results were compared with responses derived from natural day to day variation in mid-day stomatal conductance. The purpose of these experiments was to determine the magnitude of reduction in stomatal conductance at elevated [CO2], and to assess whether the relative response of conductance to elevated [CO2] was constant across light, humidity and temperature conditions. The results indicated that light, humidity and temperature all significantly affected the relative decrease in stomatal conductance at elevated [CO2]. The relative decrease in conductance with elevated [CO2] was greater at low light, low water vapour pressure difference, and high temperature in both species. For measurements made at saturating light near mid-day, the ratio of mid-day stomatal conductances at doubled [CO2] to that at ambient [CO2] ranged from 0.42 to 0.86, with a mean of 0.66 in barley, and from 0.33 to 0.80, with a mean of 0.56 in wheat. Day-to-day variation in the relative effect of elevated [CO2] on conductance was correlated with the relative stimulation of [CO2] assimilation rate and with temperature. Some limitations of multiple linear regression, multiplicative, and ‘Ball–Berry' models as summaries of the data are discussed. In barley, a better fit to the models occurred in individual years than for the combined data, and in wheat a better fit to the models occurred when data from near the end of the season were removed.