This research examines the moderating effect of auditor industry specialization on the relationship between fair value disclosure and audit fees in companies listed on the Tehran Stock Exchange. Employing a descriptive-correlation approach and an applied objective, the study utilizes panel data from 151 selected companies out of a total of 245 companies, spanning the years 1392 to 1400 (2013-2021). The sample size, calculated using Cochran's formula with a 95% confidence level, 0.5 probability of success, and a 0.05 margin of error, was selected through simple random sampling, resulting in 1359 observations. Data were extracted from library resources, the Rahavard Novin software, and the Codal website. Data validity was confirmed by comparing with official reports from the Securities and Exchange Organization and previous standard indices, while reliability was assessed using panel and time-series tests such as Levin, Lin, and Chu, all of which declared the variables as stationary. In the analysis, descriptive statistics for variables, such as the mean of the log of audit fees (12.856) and the ratio of total fair value to assets (0.143), were first calculated using EViews software. Subsequently, the Pearson correlation test ruled out multicollinearity, as the highest coefficient was 0.456. The F-Limier test confirmed the panel structure of the data, and the Hausman test indicated that a fixed-effects model was appropriate. Regression models were estimated for four hypotheses. The first hypothesis, regarding the moderation of auditor specialization based on market share on the relationship between total fair value disclosure and audit fees, was supported with an interaction coefficient of 0.912 and a probability value less than 0.001. The second hypothesis, concerning the moderation of specialization based on portfolio share on the same relationship, was also supported with an interaction coefficient of 0.945 and a probability value less than 0.001. The third hypothesis, on the moderation of market share specialization on the relationships at hierarchical levels, was supported with interaction coefficients for the first, second, and third levels of 0.789, 0.812, and 0.701, respectively, all significant with probability values less than 0.01. The fourth hypothesis, on the moderation of portfolio specialization on hierarchical levels, was also supported with interaction coefficients for the first, second, and third levels of 0.812, 0.834, and 0.723, respectively, all significant with probability values less than 0.01. Diagnostic tests, such as Jarque-Bera with a probability value greater than 0.05 and Durbin-Watson close to 2, confirmed the validity of the models with coefficients of determination ranging from 0.612 to 0.701. The results indicate that auditor specialization reduces fair value disclosure risk and positively moderates audit fees, which is applicable to stock exchange policymakers and auditors.