Portland cement production in Canada increased from 1.7 million tonnes in 1946 to more than 13 million tonnes in 2002. Although the industry is a major player in meeting infrastructural needs of Canadians, it is also a major user of natural resources such as water, natural minerals, and aggregates. Moreover, for each tonne of cement clinker produced, 1 t of CO2 is released into the atmosphere. At the same time, Canada produces nearly 5 million tonnes of fly ash each year, and yet the use of fly ash as cement replacement remains dismally low at around 17%. It is believed that through product and process innovations, more fly ash can be used in concrete, thus preserving natural resources, reducing CO2 emissions, and helping Canada meet the requirements of the Kyoto Protocol. Forecasting the future impact of using fly ash in concrete has been based on qualitative and linear estimates, however, without accounting for the complexity of the problem and its dynamic feedbacks. In this paper, a novel application of system dynamics modeling, a feedback-based object-oriented modeling paradigm, is proposed to create a rational model that departs from current approaches used in modeling CO2 emissions of cement production. The model accounts for the various enablers and barriers for using fly ash in concrete, including market dynamics and technology development. It allows the user to test a wide variety of scenarios and policies, its flexible architecture permits coupling it with general economic or service life models, and its modular nature allows expanding its boundaries to include other facets of the holistic CO2 emissions problem in Canada.Key words: concrete, blended cements, CO2 emission, global warming, sustainable development, system dynamics, modeling.