SUMMARY In three separate experiments, the response of three populations of Agrostis capillaris L., Nardus stricta L., and Lolium perenne L., from areas differing in ambient air pollution concentrations, were examined in experimental fumigations using SO 2 , O 3 , NO 2 and NH 3 . Growth of A. capillaris generally increased following the fumigations, except at the highest pollutant concentrations (120 μg m ‐3 O 3 , alone or combined with 150 μg m ‐3 SO 2 and 35 μg m ‐3 NO 2 ). Growth of each population of N. stricta was stimulated by the SO 2 and NH 3 combination, but was significantly increased only in the populations originating from the most heavily polluted of the three areas. The growth of L. perenne was not inhibited by any air pollutant or combination of pollutants. Despite a low growth rate in clean air and a large amount of within‐population variation, biomass production was significantly stimulated by the highest pollutant mixture. The response of each species to the experimental treatments depended in part on the location from which it originated. This site‐dependent (population) response differed between species, suggesting that the selective processes induced by these air pollutant concentrations were insufficient for a clear‐cut adaptation to gaseous air pollutants.