The feasibility of relating crankshaft free end torsional signals to the misfiring of an individual cylinder has been established in earlier torsional simulation work . This paper advances two diagnostic methods for detection of the IMEP level of individual cylinders.The first method is an enhanced pattern recognition method using both the flywheel torque fluctuation as the primary signal; and the model based reciprocating inertia knowledge for gas pressure torque harmonics determination. The enhanced pattern recognition method ensures IMEP level detection for each individual cylinder.The second method i s based on a reverse simulation algorithm. The dynamic flywheel speed fluctuation is used as the input signal to reconstruct cylinder pressure diagrams for each cylinder.In both methods, a rigorous, multi-mass, lumped mass, elastic crankshaft model is used to depict the mass-elasticity damping system.In the torsional vibration simulation mode, individual cylinder gas pressure and engine RPM are the two operating inputs. Flywheel torque fluctuation and flywheel speed fluctuation are the outputs.In the reverse simulation mode, the flywheel speed fluctuation is the input for reconstruction of the individual cylinder pressure. A normalized heat release rate map is used to bridge the gas pressure torque harmonics and the individual cylinder pressure harmonics.This paper reports the progress made on the analysis of the enhanced pattern recognition method and on the reverse simulation algorithm using a rigorous multi mass, multi cylinder engine model. Analysis shows that the reverse simulation algorithm is most promising for on-board firing diagnostics and cylinder pressure reconstruction.