The aim of this study was to determine the effects on respiratory drive of two factors, one mechanical (lung volume) and one chemical (sensitivity to hypercapnia), that are involved in determining the breath-hold duration (BHD). Functional residual capacity was measured by helium dilution with the subject seated in air, seated in water and in the prone position in water. Hyperoxic hypercapnia rebreathing (Read's method) was carried out under identical environmental conditions to assess the effects of CO2 pressure on respiratory centre output by measuring ventilation, mean inspiratory flow and occlusion pressure. Sixteen healthy volunteers were tested. 8 trained divers and 8 non-divers. Functional residual capacity decreased for the postures seated in water (30.8%-34.8%) and for prone position in water (20.3%-20.9%) when compared to the posture seated in air (P<0.0001), all subjects pooled. No difference was found between groups. The slopes of the linear regression, which characterised the sensitivity to CO2 and were determined with the rebreathing tests, revealed differences between the two populations (ventilation: P < 0.0001; mean inspiratory flow: P < 0.05). No difference was found for occlusion pressure or between the different postures. These results confirmed a lower sensitivity to CO2 for trained divers. This adaptation was shown to decrease respiratory centre activity at the origin of the breath-hold breaking point. The immersion, did not influence respiratory drive, despite a decrease in lung volumes. The authors suggest that these findings may be explained by a specific apnoea training and a pronounced bradycardia in immersion.