The in-cylinder trapped air, residual gas, and temperature directly impact Spark Ignition (SI) engine operation and control. However, estimation of these variables dynamically is difficult. This study proposes a dynamic cycle-by-cycle model for estimation of the in-cylinder mixture temperature at different events such as Intake Valve Closed (IVC), as well as mass of trapped air and residual gas. In-cylinder, intake and exhaust pressure traces are the primary inputs to the model. The mass of trapped residual gas is affected by valve overlap increase due to the exhaust gas backflow. Of importance to engines with Variable Valve Timing (VVT), the compressible ideal-gas flow correlations were applied to predict the exhaust gas backflow into the cylinder. Furthermore, 1D GT-Power Three Pressure Analysis (TPA) was used to calibrate and validate the designed model under steady-state conditions. To minimize the calibration efforts, Design of Experiments (DOE) analysis methodology was used. The transient behavior of the model was validated using dynamometer dynamic driving cycle. The cycle-based output parameters of the developed model are in good agreement with transient experimental data with minimal delay and overshoot. The predicted parameters follow the input dynamics propagated in the in-cylinder, intake and exhaust pressure traces with a 1.5% average relative steady-state prediction error.