TOMSIM(l.O) and TOMGRO(l.O) are two dynamic models for tomato growth and development. Their submodels for dry matter production are compared and discussed. In TOMSIM(l.O), dry matter production is simulated by a modified version of SUCROS87 (Spitters et al., 1989). Single leaf photosynthesis rates are calculated separately for shaded and sunlit leaf area at different depths in the canopy, according to the direct and diffuse components of light; daily crop gross assimilation rate (A) is computed by integration of these rates over the different depths and over the day. In TOMSIM(l.O) leaf photochemical efficiency (e) and potential leaf gross photosynthesis rate at saturating light level (Pgmax) both depend on temperature and C02 level. In TOMGRO(l.O) crop gross photosynthesis rate is calculated by the equation of Acock et al. (1978); e is a constant and Pg max is a linear function of C02. In both models leaf photosynthesis characteristics are assumed to be identical in the whole canopy. Maintenance respiration (Rm) and conversion efficiency (Cf) are taken into account in the same way, except that root maintenance respiration is neglected in TOMGRO(l.O). For both models a sensitivity analysis was performed on the input variables (light intensity, temperature, C02 and leaf area index (LAI)) and on some of the model parameters. Under most conditions considered, simulated A was found to be 5-30% higher in TOMSIM(l.O) than in TOMGRO(l.O). At temperatures above 18°C Rm was also higher in TOMSIM(l.O), and C, was 4% higher in TOMGRO(l.O). The two models were very sensitive to changes in e and to a lesser extent to changes in the light extinction coefficient, whereas the scattering coefficient of leaves had hardly any effect on the simulated A. TOMGRO(l.O) appeared to be rather sensitive to the C02 use efficiency, whereas at ambient C02 level mesophyll resistance was quite important in TOMSIM(l.O). Four sets of experimental data (differences in cultivar, C02 enrichment and planting date) from Wageningen (The Netherlands) and Montfavet (southern France) were used to validate the models. Average 24 h temperature and average daily C02 concentration values were used as input to the models. For the Wageningen experiments, hourly PAR values were calculated from the daily global radiation sum by TOMSIM(l.O) and used as input in both models. For the Montfavet experiment, average hourly PAR measurements were used. Also measured LAI, dry matter distribution and organ dry weights (for calculation of Rm) were input to the simulation. In the Wageningen experiments, total dry matter production was simulated reasonably well by both models, whereas in the Montfavet experiment an under-estimation of about 35% occurred. TOMGRO(l.O) and TOMSIM(l.O) simulated almost identical curves in all four experiments. Strong and weak points of both models are discussed.