Hypertension (HTN) as defined by the world health organization (WHO) is when the blood pressurernis elevated with a systolic blood pressure equal to or greater than 140mmHg and a diastolic bloodrnpressure equal to or greater than 90mmHg when measured on 2 separate days. HTN, according tornCenter for Disease Control and Prevention (CDC), raises the risk for stroke and other heart diseases.rnIn a previous study, it was estimated that there were 972 million adults with HTN in 2000; 333 millionrnin developed countries and 639 million people in developing countries. In a review of current trendsrnby WHO, there has been an increase in adults with HTN from 594 million in 1975 to 1.13 billion inrnyear 2015 with a prevalence of 27% in Africa and 18% in the America. Hypertensive Retinopathyrn(HPR) is damage to the retinal blood vessels caused by HTN. The use of artificial intelligence hasrnrecently been applied in Medical Imaging. One of such areas is in classification of eye diseases. InrnHPR, previous studies only focused on classification according to two classes; normal or HPR. In therncurrent study, a classification method based on four classes according to the Wong and Mitchellrnclassification which gives a classification based on the severity of HTN severity is presented; Normal,rnMild, Moderate, and Severe or Malignant HPR. The study explored the idea of applying ConvolutionalrnNeural Network (CNN) in classifying HPR. The work focused on images collected solely fromrnpatients with HTN and will play an important role in assisting physicians to speed up the process ofrndiagnosis. The study used the HPR features of the retina for classification. Transfer learning usingrnexisting pretrained models was employed in classifying the dataset. Five state-of-the-art pretrainedrnmodels were employed for classification into two classes; normal and abnormal and into four classes:rnnormal and three stages of HTR. The five different neural architectures namely: ResNet-101,rnGoogleNet, AlexNet, VGG-19 and Xception achieved an accuracy of 91.61%, 93.01%, 87.41%,rn90.21% & 85.31% for the 4 Classification task and 100%, 100%,100%, 99.30% & 97.89% for the 2rnClassification task respectively. The best accuracy was obtained using GoogleNet. The pretrainedrnmodels show that with proper tuning of training parameters like the learning rate, number of epochs,rnbatch size, type of