The prediction of recovery rate in waterflooding is becoming increasingly important as the percentage of total U. S. oil production recovered by waterflooding continues to increase. Since the first method of predicting waterflood recovery was proposed in 1949, various studies have shown that two factors effect the success or failure of a waterflood to a much greater degree than all others. These factors are, in the order of their importance, total cumulative waterflood recovery and rate of recovery. This study presents a method of predicting waterflood recovery rate based on the performance of 86 successful Oklahoma waterfloods. Introduction In 1964, secondary recovery as a percent of total statewide production was 71.2 percent in Colorado, 68.6 percent in Illinois, 32.6 percent in Oklahoma, 30 percent in Texas and 12.1 percent in Louisiana. As U. S. primary crude oil reserves found by exploration continue to decrease, secondary recovery - 30 percent of which is from waterflooding-must increasingly be a source of reserves. Fig. 1 shows actual secondary recovery as a percent of total U. S. production for the period 1950 through 1965, and the USBM projection for the period 1966 through 1980. Since secondary recovery and waterflooding are becoming increasingly more important and widespread, it becomes more pressing to find more accurate methods of predicting their performance, both in total recoverable secondary barrels and in the rate at which these barrels are recovered. The purpose of this study is to arrive at a set of empirical parameters for use in the prediction of recovery rates when waterflooding sands which, for all practical purposes, have been depleted by primary production methods. Waterflood Recovery Rate Data Monthly production and injection data were obtained for 86 successful waterfloods located in 23 Oklahoma counties. These floods represent 56 separate fields and 23 different sands (Fig. 2). Tables 1 and 2 give the basic information concerning each of the waterfloods studied. Productive areas of the floods vary from 30 to 3,050 acres. The total cumulative waterflood recovery from all floods to Jan. 1, 1967, has been 84,898,188 gross bbl with an ultimate cumulative recovery of 92,530,942 gross bbl obtained from the 2,106 producing wells. Average waterflood recovery to Jan. 1, 1967, for each of the 86 floods is 987,188 gross bbl, with an average ultimate waterflood recovery of 1,075,941 gross bbl. Total water injection (Table 2) to April 1, 1967, has been 1,231,653,314 bbl into a total of 1,892 injection wells. Flood Life Periods Waterfloods can be divided into the following distinct periods. Initial Response-the period from initial water injection until the first response to injection in the form of an oil production increase. During this period the oil production rate may decline or remain constant. The reservoir voids are filled, free gas goes back into solution, and the reservoir pressure is restored. Of the successful waterfloods studied, the percentage of the total flood life in this period varied from 0.5 to 21.2 percent. JPT P. 933ˆ