This study develops a transfer learning model for the automated classification of two species of fruit flies, Anastrepha fraterculus and Ceratitis capitata, in a controlled laboratory environment. Accurate identification of these species is crucial for pest control and agricultural management, but current methods rely on manual classification by experts, which is time-consuming and prone to human error. This research addresses these limitations by capturing high-quality images using a mobile phone camera and a stereo microscope, followed by segmentation to focus on relevant morphological features. The images were labeled and preprocessed to create a consistent dataset for training the pre-trained convolutional neural network models VGG16, VGG19, and Inception-v3. The results evaluated using the F1-score, showed 82% for VGG16 and VGG19, while Inception-v3 achieved 93%. Inception-v3’s reliability was confirmed through testing in uncontrolled environments, supported by Grad-CAM analysis, demonstrating its potential for accurate and automated classification of fruit fly species in real-world applications.