The computer technique explained here, which is useful in analyzing complex transient test data is illustrated by two field examples. In the first example the method is used to analyze data from a multiple-well interference test in an anisotropic reservoir. In the second, it is used to examine a pressure fall-off test with significant wellbore storage. Introduction Pressure transient testing can provide better average Pressure transient testing can provide better average values of in-situ reservoir permeability and porosity than core analysis or logging techniques. It can also be used to provide information about directional permeability, barriers to fluid flow, and wellbore permeability, barriers to fluid flow, and wellbore damage or stimulation. Data analysis methods include graphicals and automated techniques. The graphical techniques have the disadvantages of being restricted to relatively simple cases and of requiring subjective judgment. The automated techniques (regression and least-squares linear programming) eliminate this subjectivity. Because the automated techniques rely on digital computers, they can be used to study complicated systems and handle large quantities of data, especially data from several wells with varying flow rates. In this paper, we describe an automatic transient-test analysis technique that utilizes regression theory and the exponential integral solution to the diffusivity equation. This approach allows computer analysis of complex transient test data from systems with varying flow rates and with directional permeability. Pressure buildup, fall-off, and drawdown tests, injectivity tests, interference tests, or a combination of these tests may be analyzed. Data may be analyzed for individual wells, or simultaneously with data from other wells. To illustrate the use and value of the technique, we present the results of applying the approach to two field present the results of applying the approach to two field cases. The first field case study shows how the method can be used to analyze data from a multiple-well interference test in an anisotropic reservoir. Pressure data are from six observation wells surrounding an injection well that experienced a sporadic injection history. The irregular injection schedule would have made conventional manual data analysis extremely difficult. Automatic analysis yielded permeability values, the direction of permeability anisotropy, and an indication of reservoir permeability anisotropy, and an indication of reservoir storage capacity. These results are supported by field waterflood performance and by oriented-core data. The second case study examines a pressure fall-off test with significant wellbore storage. By accounting for the varying sand-face flow rate, automatic analysis provides a good method for calculating permeability. provides a good method for calculating permeability. This test included additional injection and shut-in periods, with pressure measured not only in the periods, with pressure measured not only in the injection well, but also in an observation well. Results from analysis of the observation-well data verify the analysis of the wellbore-storage-dominated fall-off data. Analysis Method Least-squares theory provides a rational method for selecting the parameters in any equation that result in a best fit of some observed data. JPT P. 1271