ARI (Acute Respiratory Infection) is a common disease in the community, especially infants and children caused by rhinoviruses, RSV, adenovirus, influenza virus or parainfluenza virus. Most people do not know what are the symptoms arising from ARI suffered by children. Pediatricians have data on the symptoms caused by ARI, with which the data can identify the child's disease. However, the public must go to the doctor first to find out what ARI diseases are suffered by children based on the symptoms felt. Checking the symptoms of ARI disease is diagnosed directly by the medical party, namely the doctor by recording and analyzing manually. With conditions like this, of course, there are many obstacles that do not even rule out the possibility of different diagnoses of diseases experienced by patients. The expert system that will be built is intended to make it easier for someone to diagnose ARI disease by adopting human knowledge into the computer system so that the expert system is able to solve existing problems. This study aims to diagnose ARI in children using the certainty factor method based on the symptoms felt and find a solution or obtain confidence in the percentage of the disease suffered. The expert system is created by utilizing PHP and MySQL programming languages as databases. Data collection was carried out starting in August 2022, followed by data processing and analysis. The sample in this study was patients with ARI disease in children at RSU Imelda Pekerja Indonesia Medan. The sampling technique in this study is purpose sampling. Statistical analysis used one sample T-test. The targeted output is scientific publication in nationally accredited journals. TKT research proposed is as a technological research or experiment from existing theories in diagnosing ARI disease in children based on websites.