In this article, a new approach for reducing the model of a higher-order system is proposed. The method presented is a mixture of two mathematical methodologies for reducing the high order models. In the given process, stability equation method (SEM) is first used in both numerator and denominator for a certain order and then differentiation method (DM) is applied to reduce it to the desired lower order system. The efficiency and accuracy of this technique is illustrated by different numerical examples. The important benefit of the proposed technique is that it ensures the stability in the low order system and necessary features of the large-scale system. The superiority of the proposed method is observed by comparison of the responses of the both the higher-order and lower-order systems. The performance of the higher-order model and reduced-order model is also compared in terms of error indices like integral time absolute error (ITAE), integral absolute error (IAE), and integral square error (ISE). Finally, the effect of time delay on the closed-loop model of the original and reduced-order model is verified.