Time series referring to water prices at different regions all over the world are examined in this paper by using fractionally integrated methods. We look at series corresponding to the following regions: Asia Pacific and Russia, Europe, United States and Latin America as well as global data. The results indicate large degrees of persistence, with the values of the differencing parameter being close to one in all cases and higher under the assumption of uncorrelated errors. If autocorrelation is permitted, a small degree of mean reversion is found in all except the Latin American series. The possibility of structural breaks is also investigated and the results indicate the presence of multiple breaks in the data: three in the case of Latin America and global data; four in Europe and USA and five for the Asian Pacific and Russia. Nevertheless, we do not observe a significant change in the degree of persistence across subsamples and once more mean reversion is found if autocorrelation is permitted.