In this study, the effect of COVID-19 on the private education participation rate (PEP rate) by type of tutoring and income class was analyzed using data from the Gyeonggi Education Panel Study (GEPS). The double-differentiation (difference in difference; DID) method was used to remove the grade effect from the longitudinal data. The results showed that when the grade effect was removed, the overall PEP rate decreased by 15.96%p due to the COVID-19 situation in 2020. The participation rates in one-to-one/group tutoring and private institutes decreased by 4.42%p and 9.05%p, respectively, but the participation rate in workbook/online tutoring increased by 1.06%p compared to 2019. When the grade effect was removed, the overall PEP rate in all income classes, except group 5, decreased due to the impact of COVID-19 in 2020. With the grade effect removed, the PEP rate in one-to-one/group tutoring and private institutes decreased in all income groups in 2020, but the participation rate for one-to-one/group tutoring greatly decreased in the high-income class, and that for private institutes decreased significantly in the low-income class. On the other hand, the participation rate for workbook/online tutoring increased in the low-income class and decreased slightly in the high-income class. As the overall inequality of educational opportunities intensifies due to the COVID-19 situation, and the pattern of changes in inequality tends to differ depending on the type of private education, appropriate policy input is needed to meet the needs of the underprivileged.