Studies on lung injury and its treatment options are often performed on small animals like rats. Because conventional blood gas analyses may not detect rapid changes in gas exchange during respiratory distress syndrome and intermittent blood withdrawal can result in hypo-volaemia and anaemia, we tested the applicability and accuracy of a continuous intravascular blood gas monitor (Paratrend 7+). Anaesthetized and ventilated rats with a body weight of 398 ±45 g (n =22) had a 20-gauge cannula inserted in both carotid arteries. A photochemical blood gas sensor for continuous measurement (Paratrend 7+) was advanced into the aorta via the left carotid artery. Blood was sampled for intermittent blood gas analysis by means of the right carotid artery. Arterial pO₂ was varied by applying different inspiratory oxygen concentrations, and arterial pCO₂ by applying different respiratory rates. Paired blood gas measurements (n =136) were analysed over a wide range of pO₂ values (5.3-76.8 kPa). We found an acceptable correlation for pO₂ (r ²=0.98), pCO₂ (r ²=0.96) and pH (r ²=0.92). The calculated bias and imprecision for pO₂ was -1.0 ± 3.3 kPa, for pCO₂ 0.04 ± 0.28 kPa and for hydrogen ion concentration -0.05 ±2.2 nmol/l. We conclude that in rats, continuous blood gas monitoring with a photochemical blood gas sensor provides pO₂, pCO₂ and pH measurements with acceptable accuracy.