Architecture has evolved in different styles expressing the culture, aesthetic values, technology, and other characteristics of the eras in which it was born. Numerous studies including history, restoration, and architectural design theory are interested in the classification of these styles. However, this process is crucial in traditional methods. Recently, significant advancements have been made in this domain through the implementation of computer-based methods, utilizing deep learning (DL) and convolutional neural networks (CNNs). Several studies have employed these techniques, but challenges such as limited large training datasets and overfitting in models persist. To address these issues, this study employs Transfer Learning for the classification of architectural styles. Multiple models were tested to achieve the highest accuracy and best performance, using an open source dataset named ’Architectural Styles.’ Finally, the results of different models are presented according to the selected evaluation metrics.