The present work describes an original method to follow rate of (CO2)-C-14 and total CO2 production from rhizosphere respiration after plant shoots had been pulse-labelled with (CO2)-C-14. We used a radioactivity detector equipped with a plastic cell for flow detection of beta radiation by solid scintillation counting. The radioactivity detector was coupled with an infrared gas analyser. The flow detection of (CO2)-C-14 was compared to trapping of (CO2)-C-14 in NaOH and counting by liquid scintillation. First, we demonstrated that NaOH (1 M) trapped 95% of the CO2 of a gaseous sample. Then, we determined that the counting efficiency of the radioactivity flow cell was 41% of the activity of gaseous samples as determined by trapping in NaOH (1 M) and by counting by static liquid scintillation. The sensitivity of the (CO2)-C-14- flow detection was 0.08 Bq mL(-1) air and the precision was 2.9% of the activity measured compared to 0.9% for NaOH trapping method. We presented two applications which illustrate the relevance of (CO2)-C-14-flow detection to investigations using C-14 to trace photoassimilates within the plant-soil system. First, we examined the kinetics of (CO2)-C-14 production when concentrated acid is added to (NaHCO3)-C-14. This method is the most commonly used to label photoassimilates with C-14. Then, we monitored (CO2)-C-14 activity in rhizosphere respiration of 5-week old maize cultivated in soil and whose shoots had been pulse-labelled with (CO2)-C-14. We conclude that alkali traps should be used for a cumulative determination of (CO2)-C-14 because they are cheap and accurate. On the other hand, we demonstrated that the flow detection of (CO2)-C-14 had a finer temporal resolution and was consequently a relevant tool to study C dynamics in the rhizosphere at a short time scale.