For most studies involving the response of plants to future concentrations of atmospheric carbon dioxide (CO2), a current concentration of 360–370 μatm is assumed, based on recent data obtained from the Mauna Loa observatory. In the present study, average seasonal diurnal values of ambient CO2 obtained at ground level from three global locations (Australia, Japan and the USA) indicated that the average CO2 (at canopy height) can vary from over 500 μatm at night to 350 μatm during the day with average 24-h values ranging from 390 to 465 μatm. At all sites sampled, ambient CO2 rose to a maximum value during the pre-dawn period (03.00–06.00 hours); at sunrise, CO2 remained elevated for several hours before declining to a steady-state concentration between 350 and 400 μatm by mid-morning (08.00–10.00 hours). Responses of plant growth to simulations of the observed variation of in situ CO2 were compared to growth at a constant CO2 concentration in controlled environment chambers. Three diurnal patterns were used (constant 370 μatm CO2, constant 370 during the day (07.00–19.00 hours), high CO2 (500 μatm) at night; or, high CO2 (500 μatm) at night and during the early morning (07.00–09.00 hours) decreasing to 370 μatm by 10.00 hours). Three plant species − soybean (Glycine max, L (Merr.), velvetleaf (Abutilon theophrasti L.) and tomato (Lycopersicon esculentum L.) − were grown in each of these environments. For soybean, high night-time CO2 resulted in a significant increase in net assimilation rate (NAR), plant growth, leaf area and biomass relative to a constant ambient value of CO2 by 29 days after sowing. Significant increases in NAR for all three species, and significant increases in leaf area, growth and total biomass for two of the three C3 species tested (velvetleaf and soybean) were also observed after 29 days post sowing for the high night/early morning diurnal pattern of CO2. Data from these experiments suggest that the ambient CO2 concentration experienced by some plants is higher than the Mauna Loa average, and that growth of some agricultural species at in situ CO2 levels can differ significantly from the constant CO2 value used as a control in many CO2 experiments. This suggests that a reassessment of control conditions used to quantify the response of plants to future, elevated CO2 may be required.