Learning disabilities are generally defined as disorders of one or more basic psychological processes including the use of spoken or written language shown by children with poor ability to hear, think, read, write, spell, or solve mathematic problems. The purpose of this research is to classify learning disabilities in children using a fuzzy expert system as a step in recognizing the learning abilities of school-age children so that treatment can be done early on. This expert system is built with several modules, namely a knowledge base, memory, fuzzy inference engine, and a web-based interface. Answers from users will be processed using a fuzzy expert system based on the scores obtained from each category of reading, writing, math and logic, and heredity. The output of this system answers whether the child is normal, dyslexia, dysgraphia, dyscalculia, or a combination of the types of learning disabilities. The expert system was conducted on 20 school-age children (6-12 years old). The classification results of three types of learning disabilities using a fuzzy expert system showed an accuracy of 95%.