Eye diseases, especially those affecting the retina, pose significant challenges in the field of medical diagnostics. Accurate and timely classification of these diseases using fundus images is essential for effective treatment. This research paper explores a novel approach to enhance the classification of eye diseases by leveraging the power of hybrid classical-quantum machine learning techniques. In this study, we employ classical transfer learning models with state-of-the-art convolution neural network, ResNet18, ResNet50 and ResNet101, to extract informative features from fundus images. These features provide valuable insights into the underlying pathology of the eye, yet traditional classification methods often face limitations in handling the complexity and intricacy of such data. To address these challenges, we integrate quantum computing into the classification pipeline. Quantum circuits are employed to fine-tune the classifier, taking advantage of quantum parallelism and optimization capabilities to optimize the decision boundaries. The hybrid approach synergistically combines the strengths of classical and quantum computing, resulting in a more robust model along with reduced learning time for multi-class classification. The dataset used in this paper is taken from kaggle and contains the fundus images of four classes namely - cataract, diabetic retinopathy, glaucoma and normal. The simulation are carried using python library pennylane. Our experimental results demonstrate the potential of this hybrid approach in achieving good classification accuracy for diabetic retinopathy and cataract.