Candidiasis is an infectious disease caused by the fungus candida. Research on this fungus has been widely carried out until several types of candida fungi are found that can attack and cause infections in humans. Types of candidiasis also vary, but can be classified in general into three types, namely attacking the mouth (Candidiasis Thrush), vagina (Vulvoginal Candidiasis), and skin (Cutaneous Candidiasis). Candidiasis is very susceptible to infection and infection, therefore a study is needed to diagnose candidiasis. Today, expert systems are often used to diagnose diseases. There are several methods commonly used in expertise, including the Certainty Factor method and the Bayes Theorem. However, the problem faced in implementing an expert system in any field is uncertainty. This is caused by the user's hesitation in answering questions during the consultation session or even the inaccuracy of the methods used in building the system. Therefore, it is necessary to study and compare the methods that can be used to build the system. Exponential is a simple comparison that can reduce bias in the analysis process. This study aims to apply and analyze both methods and the results compare with an exponential comparison in detecting candidiasis in humans. The results of this study showed that both methods achieved the same results, namely the lowest percentage level was Candidiasis Truth, then Vuvoginal Candidiasis, and the highest was Candidiasis Cutaneous. Of these two methods, Certanty Factor is more accurate in diagnosing candidiasis.