In this paper we try to measure oil price uncertainty. The measure of uncertainty is based on the conditional standard deviations which are derived from univariate (G)ARCH models. The measure of uncertainty we choose is the within-year high-low range of the conditional standard deviations. It is likely that the higher the uncertainty, the higher the high-low range within a year will be. We focus on volatility of the price of a barrel Brent crude, over the period 5 January, 1982 to 23 April, 2002 representing 5296 observations. The preferred model is a symmetric GARCH(1,3) model. Asymmetric leverage effects are not found. We also examine the volatility in monthly time series for the period January, 1970 to April, 2002. For this time span and frequency we prefer the GARCH(1,1) model.