Improvements in forecast models have reduced track errors, but intensity forecasts have not shown significant gains. Numerical models dealing with intensity change are not much better than CLIPER models, and statistical models do not address rapid deepening. Intensity studies have focused on three areas: air-sea interaction, internal dynamics, and external (environmental) interactions. Since the work of Hoskins et al. (1985), potential vorticity (PV) has gained popularity as a research tool for tropical cyclones. PV has been used to understand internal and external influences on tropical cyclone formation and intensity. Little is known about the causes of intensity changes of recurving typhoons. This study will use PV to examine external influences on the intensity of recurving typhoons. Typhoon track and intensity data were obtained from Annual Tropical Cyclone Reports (Joint Typhoon Warning Center). PV was calculated on isentropic surfaces from 310K to 370K with reanalysis data obtained from the European Centre for Medium-Range Weather Forecasts. The data set has a horizontal resolution of 1.125 X 1.125 degrees and a vertical resolution of 13 pressure levels. From 1988-1997, 71 typhoons recurved. Of these storms, 31 intensified, 28 weakened, and 18 do not change intensity. Recurving time in this study is defined to be when the storm heading is between 315 degrees (where the storm will be moving more north of west) and 45 degrees (where the storm will be moving more east of north). Typhoons recurve mainly from a trough interaction or from a weakening in the subtropical ridge. This study will look at three different intensity classes of recurving typhoons; those that intensify, those that weaken, and those whose intensity remains constant. The Dvorak method is used to determine these intensity changes. PV analysis at upper levels has been useful in examining trough interactions, but little work has been done using PV without a trough interaction. This analysis is going to look at both situations. One example of a case with a weak ridge was Super Typhoon Ginger in 1997. Ginger intensified from 40 kts. to 145 kts. during recurvature. There was not a pronounced PV gradient in the upper levels. This is to be expected as PV is highest in the tropopause and stratosphere. Without a trough present, there is not a pronounced decrease in tropopause height. As the storm intensified, low values of PV air were advected outward from the storm. In the middle troposphere, there were some interesting PV anomalies. A high PV anomaly formed on the southeastern side of the storm, and tracked cyclonically around the storm as it intensified. Our preliminary interpretation is that the high PV anomaly was due to diabatic heating from a convective outbreak and may have been favorable for storm intensification. After the storm reached maximum intensity, the high PV anomaly settled over the storm, and expanded outward. We believe this behavior may indicate winds in the outer part of the storm continued to increase even though winds in the eyewall decreased. At the conference, a comparison of intensifying, weakening, and steady storms will be made to determine if there is a repeatable PV signature associated with each group. Ultimately, our goal is to determine if PV fields alone can be used to forecast intensity, or do we need to use it in conjunction with other fields?