Hi Julie,
Percent Slope ranges in value from 0% to infinity. AKA Undefined Slope. Undefined slope would typically ocurr where a vertical cliff is. Your rise is positive, but your run is 0. So the equation rise/run * 100 would result in undefined slope. If your rise is 2 and your run is 1, then 2/1*100 is 200% slope. Slope is not limited to 0-90, unless you are choosing the degree slope option rather than percent.
Quite often I see percent slope values in the millions. This ocurrs when the data has different z units than it does for x and y. i.e. the data is in a geographic coordinate system using angular units like decimal degrees, but the z unit is meters/feet. An example would be a rise of 1 and a run of 1 pixel (in GCS units) * 100. 1/0.00083 * 100 = ~ 120,000% slope. You can see how not having the same units can lead to somewhat misleading slope values rather quickly. In such cases one would project the data so the x, y, and z have the same units, or they could set an appropriate z factor in the tool.
I see two scenarios for you. One, the data is projected in state plane with units of feet at a 2ft resolution and your pixel values (z's) are also in feet. This is what you have stated the situation is. You have rugged terrain, something near vertical, or at least one area that is very steep in the data resulting in percent slope as high as 6881%. Your result is correct. Two, you have x and y in one unit, but unknown to you the z unit is actually in different. This situation of mixing units feet/meters is resulting in abnormally high % slope values. Your analysis should be run again using the correct z factor depending on what unit you need to convert to.
Hope this helps you out!
