Abstract: The El Niño-Southern oscillation (ENSO) is the strongest interannual oscillation in the Earth's climate system, which has a significant impact on global weather and climate. The Holocene is the geological time most closely related to humans, and reconstructing the activity history of ENSO and exploring the change pattern of ENSO in this period will help to improve the accuracy of future ENSO prediction. In this regard, based on coral, a high-resolution climate record carrier of tropical oceans, we firstly introduce the method of extracting ENSO signal from coral skeleton index and measuring ENSO variability; then compares the history of ENSO variability in early, middle and late Holocene; and finally summarizes the mechanism of ENSO variability in different time scales in Holocene based on coral record. The results show that the ENSO signals in the coral records can be directly identified from the extreme changes of their environmental proxies; or the ENSO cycles of the environmental proxies on the time series can be extracted by spectral analysis and filtering, and then the frequency and amplitude changes of ENSO can be quantitatively analyzed by using threshold analysis and sliding window methods. The coral records show that the Holocene ENSO is characterized by fluctuating changes, with a general trend of decreasing ENSO variability from the early to middle Holocene and increasing ENSO variability in the late Holocene. Based on the coral record, it is concluded that the change in surface solar radiation distribution due to the age difference is the main factor of the century-millennium scale ENSO variation in the Holocene, while the internal drive of the climate system may be the main reason for the interannual-interdecadal scale ENSO fluctuation in the Holocene. However, compared with the long time span of the Holocene, the accumulated time window of the coral record is only a few hundred years, which is far from revealing the patterns and mechanisms of ENSO activity. Therefore, it is necessary to further extend the time series and increase the spatial area of the coral record in the future to reveal the patterns and mechanisms of ENSO variability.