We aimed to study the dynamics of ecological processes concerning primary production of an irregularly flooded short Spartina alterniflora (height 49.98 ± 20.65cm) salt marsh in the estuarine zone of Patos Lagoon (Rio Grande do Sul, Brasil, 32º10’S and 52º15’W), using a simulation model as a tool. The state variables were live aboveground biomass of Spartina, standing dead biomass, detritus from Spartina at the sediment’s surface, live and dead belowground biomass. Simulated processes were primary production, mortality, allocation, reallocation, respiration and decomposition. Forcing functions were light radiation, temperature, salinity, water level, precipitation, air humidity and percentage of nitrogen in the aerial live vegetal tissues. The model was simulated for September/1992 to May/1994. In order to check the model, Spartina were monthly collected for this period, sorted, dried at 80o C and weighted. Abiotic data were daily measured at the study site. Shoots (live and dead) showed a strong seasonal pattern, ranging from 798.85 ± 172.66 g m-2 to 304.12 ± 55.53 g m-2 . Roots varied from 3977.07 ± 687.40 g m-2 to 1477.47 ± 551.49 g m-2 , but without seasonal pattern. The model showed a good agreement with observed data, especially for live above and belowground biomass. The sensitivity analysis indicated temperature as the main system controller. Changes in temperature values modified not only biomasses values of all state variables, but also changed the seasonal patterns of standing dead and dead belowground biomasses. Primary production and translocation processes were very sensitive to environmental changes. The use of a function representing losses of detritus from the marsh surface avoided a detritus accumulation effect in the simulation, which represented an amount of 573 g m-2 of organic matter to be consumed by marsh organisms or exported to the estuary by tide or runoff. This value was similar to annual aerial productivity estimated to these marshes. The estimated aerial productivity of the model (673.69 g m-2 year-1 ) was very similar to that estimated by Smalley method (668.79 g m-2 year-1 ), as well as the turnover rate, with values of 2.93 and 3.08 year-1 , respectively. This model allowed us to evaluate the influence of abiotic factors on primary productivity of Spartina marshes in Patos Lagoon, and indicates temperature as the most important forcing function to productive process. It also allowed us to estimate the amount of litter which left the marsh, been exported or consumed during this study, with the rain playing a major role in this process.