Accurate measurements of carbon dioxide (CO2) flux from soil are important because this flux is an important component of the surface carbon budget, and a good indicator of the level of soil microbial and root activity. Half-hour CO2 fluxes from bare soil were measured using soil chamber and Bowen ratio/ energy balance (BREB) methods for 4 days in December 1992, at the Blackland Research Center, Temple, TX. Soil chamber CO2 measurements were made sequentially at nine positions in the field. Three CO2 BREB systems were used. The CO2 flux was approximate to 0 in the early morning and after sunset and was maximum (slightly less than 0.1 mg m(-2) s(-1) (2.3 mu mol m(-2) s(-1))) near midday. The coefficient of variation (CV) of chamber CO2 fluxes across the nine positions averaged 40% throughout the day, indicating the need for a large number of chamber measurements to obtain a representative CO2 flux measurement. The CV of the three daily BREB CO2 fluxes was less than 2%, indicating the BREB CO2 fluxes from the three systems were equal. There was good agreement between fluxes from the two methods. The average chamber and BREB CO2 fluxes for the entire period of measurements were 0.039 and 0.042 mg m(-2) s(-1), respectively, while the root mean square difference between half-hour fluxes from the two methods was 0.017 mg m(-2) s(-1). The methods are complementary and can both be used for soil CO2 flux measurements. The chamber method is low cost and easy to use, and offers the possibility of replicated measurements over space. The BREB method integrates over a large spatial area and is thus less affected by the high spatial variability of soil CO2 flux.