Abstract: If simulation models are written based on first principles in a declarative equation based modeling language, such as Modelica, their simulation often consists in solving a system of differential algebraic equations(DAEs). For complex physical plants, the resultant DAE system is of very large dimension, so its numerical solution requires highly intensive computation. In this paper, techniques for the symbolic manipulation of general DAE systems are presented, and used for model reduction purpose, to support efficient simulation of complex multi-domain continuous systems. The specific problems addressed are canonical transformation of equations, efficient elimination of trivial equations by means of substitution to reduce the dimension of the DAE system, and partitioning the overall system into subsystems which can be solved in sequence. The result achieved by applying the discussed algorithms on an oscillator system is illustrated. The proposed strategies and algorithms have been successfully implemented in a modeling and simulation system, named EMWorks.