Automatic age classification has drawn the interest of many scholars in the fields of machine learning and deep learning. In this study, we looked at the problem of estimating age groups using different biometric modalities of human beings. We looked at the problem of determining age groups in humans using various biometric modalities. Specifically, we focused on the use of transfer learning for sclera age group classification. 2000 Sclera images were collected from 250 individuals of various ages, and Otsu thresholding was used to segment the images using morphological processes. Experiment was conducted to determine how accurately the age group of a person can be classified from sclera images using pretrained CNN architectures. The segmented images were trained and tested on four different pre-trained models (VGG16, ResNet50, MobileNetV2, EffcientNet-B1), which were compared based on different performance metrics in which ResNet-50 was shown to outperform the others, resulting in an accuracy, precision, recall and F1-score of 95% while VGG-16, EffcientNetB1, and MobileNetV2 had 94%, 93%, and 91%, respectively. The findings from the study showed that there is an aging template in the sclera that can be utilized to classify age.