Abstarct: In today's world of economics, one of the most important and debated topics in digital currencies is their high volatility. The main purpose of this study is to model bitcoin volatility. For this purpose, bitcoin prices in the period 2016 to 2020 have been extracted from CoinMarketcap.com as data of this research. To test the research hypothesis, ARCH, GARCH and several GARCH family tests were applied to the data. The predictive power of GARCH and its family models has also been measured for the data. In the results, the coefficients obtained from the application of ARCH and GARCH models are quite significant, which confirms that the data follow these models. For GARCH family models, the significance and sign of the coefficients indicate the leverage effect and the conditional mean effect. Comparison of forecast evaluation indices also indicates that the ability of all these models to predict variance is almost the same