Chronic kidney disease (CKD) ranks among the top 20 causes of death worldwide, affecting approximately 10% of adults. CKD impairs normal kidney function. The rising incidence of CKD underscores the need for effective prophylactic measures and early diagnosis. A novel aspect of this research is the development of a technique for diagnosing chronic renal diseases. This work aids researchers in exploring early detection methods for CKD prevention using deep learning (DL) techniques. The study involved the construction of a model for CKD diagnosis based on deep neural networks (DNN). A dataset of 400 patients with 24 features was analyzed, with mean and mode statistical analysis techniques employed to substitute missing numerical and nominal values. The effectiveness of the DNN has been demonstrated in the diagnosis results, achieving an accuracy of 98.33%. DL models employ complex algorithms to analyze large datasets containing various patient information, such as age, gender, lifestyle habits, and medical history. By automatically analyzing these factors together, the model can identify patterns indicative of potential kidney issues earlier than traditional methods. In addition to more accurate CKD prediction, DL models provide faster results, potentially leading to earlier interventions or treatments by physicians.