The effective classification of lung diseases is crucial for diagnosing various conditions, requiring the accurate identification of multiple disease types from chest X-ray images. This research investigates transfer learning techniques for multi-class lung disease classification using pre-trained convolutional neural network models to enhance performance. Among the tested models, DenseNet201 demonstrated the highest accuracy of 97.43%, showcasing its ability to learn features through densely connected layers. The results highlight the efficacy of transfer learning in creating reliable and efficient solutions for lung disease classification and enhancing diagnostic accuracy. This work emphasizes utilizing advanced techniques to handle medical imaging and healthcare challenges.