Nowadays for the management of environmental issues computerized tools, particularly mathematical modelling are used mostly. System dynamics analysis issues in which time is an important factor and involve investigation of how to react and defend a system against shocks that would be applied from outside. In addition, the modelling can be used as an efficient tool in design and optimization of constructed wetlands. There are several methods for modelling of constructed wetlands such as simple methods like black box models and complex models like process models. Comprehensive and complex models improvements created this wrong belief that complex models have better performance than simple models certainly however, the complexity of the model, reliability requirements, operation’s cost and time must be taken for appropriate model. In this research removal of organic pollution, nitrogen and phosphorous in constructed wetland was simulated with STELLA software and by system dynamics method. For this research horizontal flow and subsurface flow constructed wetlands was used. was used for simulation of organic and nitrogen pollution as a case study was used for simulation of organic and nitrogen pollution as a case study. Another one was used for simulation of phosphorous. For organic matter’s modelling, chemical oxygen demand (COD) was considered as a qualitative parameter which shows organic matters. For nitrogen model Nitrate nitrogen, ammonia nitrogen and organic nitrogen and for phosphorous model total phosphorous were investigated as parameters. Processes considered for removal of BOD are included microbial degrade, uptake plant and settlement and for removal of nitrogen are included settlement, uptake plant and denitrificatiohenomenon. In addition, processes considered for removal of phosphorous are included bed’s adsorption, settlement and plant uptake. In the first case study for calibration and validation of created models from twenty four data Which determined every fifteen days, sixteen data for calibration period and eight data for validation period were used. In the second case study from sixteen data which determined every fourteen days, eight data for calibration period, and eight data for validation period were used. and eight data for validation period were used. In the BOD model from total BOD removal that was 68.7%, microbial degrade share was 29.8%, lant uptake share was 18.5% and settlement of dissolved organic matter was 20.4%. Moreover, the results for nitrogen models showed from total nitrogen removal that was 36.2%, organic nitrogen settlement share was 7.8%, ammonia nitrogen adsorption share was 7.0%, denitrification share was 16.2% and nitrate nitrogen adsorption by plant share was 5.0%. In addition, from total phosphorous removal that was 13.2%, Bed’s adsorption share was 5.6%, plant uptake share was 5.3%, and settlement share was 2.2%. To appraisal the model’s performance and constructed wetland’s behavior prediction under various conditions, four scenarios with different surface loads Were considered and the results showed removal efficiency in phosphorous model had most dependency to surface load than other models. Key Words: Constructed Wetland, Wastewater Treatment, System Dynamics, Nitrogen, Phosphorous, BOD, Modelling.