A new approach for identification of multi-input multi-output (MIMO) continuous-time transfer function models using simultaneous step input signals is proposed. MIMO processes exhibit directionality Which implies that the output gains depend on the input magnitudes as well as the ratios of the inputs. Due to this :directionality issue, MIMO: models estimated by sequentially changing one input at a time often do not result in satisfactory tracking, performance When used for model based control. In the proposed methodology all of the inputs are changed Simultaneously to resemble controlled conditions. While the identification tests remain multi-input in its true sense, the parameter estimation steps involve estimation of the parameters of a single transfer function at a time. Moreover, the time delays are estimated in the same way as the coefficients in the model. Simulation results are presented to demonstrate the robustness of the methodology and its applicability to processes with high input output dimensions. Identification and control results of a simulated distillation column are also presented; a dynamic matrix Controller (DMC) results in better control Performance with the model estimated using the proposed methodology than with the model estimated using sequential inputs.