Lung cancer continues to be one of the most lethal cancers globally, with early and accurate diagnosis being pivotal for improving patient outcomes. This study investigates the effectiveness of three DenseNet architectures namely DenseNet121, DenseNet169, and DenseNet201 in the classification of lung cancer using histopathologic images from the LC25000 dataset, with a specific focus on 15,000 lung images. Comprehensive evaluations were conducted to compare the performance of these models. The results reveal that DenseNet201 achieves superior performance with an accuracy of 99.23%, surpassing DenseNet169 and DenseNet121. This high level of accuracy underscores the potential of DenseNet201 for integration into clinical workflows, offering a robust tool for the early detection and diagnosis of lung cancer. Our findings suggest that deeper DenseNet architectures are particularly well-suited for this task, providing a significant advancement in the use of deep learning for medical image analysis.