Some studies of responses of plants to elevated concentrations of carbon dioxide (EC) added CO2 only in the daytime, while others supplied CO2 continuously. I tested whether these two methods of EC treatments produced differences in the seed yield of soybeans. Tests were conducted for four growing seasons, using open top chambers, with soybeans rooted in the ground in field plots. One third of the chambers were flushed with air at the current ambient [CO2] (AC), one third had [CO2] 350 µmol mol-1 above ambient during the daytime (ECd), while one third had [CO2] 350 µmol mol-1 above ambient for 24 h per day (ECdn). ECdn increased seed yield by an average of 62 % over the four years compared with the AC treatment, while ECd increased seed yield by 34 %. Higher seed yield for ECdn compared with ECd occurred each year. In comparing years, the relative yield disadvantage of ECd decreased with increasing overall seed yield. On days with high water vapor pressure deficits, soybean canopies with ECd had smaller midday extinction coefficients for photosynthetically active radiation than canopies with ECdn, because of a more vertical leaf orientation. Hence the seed yield of soybean at EC varied depending on whether EC was also provided at night, with much greater yield stimulation for ECdn than for ECd in some years.