From photosynthetic studies on a range of monocotyledonous (C-3 and C-4) and dicotyledonous (C-3) plants using a leaf oxygen electrode, we conclude the following. (i) A non-linear model [J.H.M. Thornley (1976) Mathematical models in plant physiology, Academic Press, London; B. Marshall and P.V. Biscoe (1980) J Exp Bot 31:29-39] significantly better describes the photosynthetic light response curve [rate of photosynthesis (P) versus incident photosynthetic photon flux density (I)] than the frequently used linear hyperbolic model [E.I. Rabinowich (1951) Photosynthesis and related processes, vol 2, Wiley, New York]. (ii) When used at the recommended CO2 partial pressures (Ca = 1-5 kPa), CO2 supply saturates the photosynthesis rate in the C-3 dicot Phaseolus coccineus L. but not in the C-3 monocot Hordeum vulgare L.. (iii) Fits using a linear hyperbolic model for P versus I produce relatively large and statistically significant errors (approximately 60%) in the estimation of Pmax and quantum efficiency (alpha) if Ca is not > 5 kPa. (iv) The convexity term, theta, incorporated into the non-linear models for P versus I appears to reflect the limitation placed on the carboxylation processes by the supply of CO2 to the chloroplast stroma. Therefore, the use of a non-linear model providing an estimate of theta should be encouraged, as it is likely to provide information on the physiological status of plants.