This paper addresses important concerns in using statistical data to track outcomes of people with disabilities and provides new evidence of employment trends of people with disabilities using alternative disability conceptualizations from the Survey of Income and Program Participation (SIPP). This analysis comes at an important time because some researchers have criticized the data and definitions used to measure these trends. At the extreme, some have concluded that such analyses should cease because of major limitations in measuring disability that exists in current surveys (especially the SIPP). Because the SIPP has been used extensively to examine outcomes of people with disabilities, it is important to understand these data criticisms and test whether the trends from the SIPP mirror those in other data sources. We conclude that the different empirical results found by researchers are not caused by "problems" with the data but rather with the assumptions researchers make when using the data. We illustrate the importance of exercising caution when developing disability questions and measuring disability trends in existing data sources. While some measures of limitations may be problematic, we find that the relatively broad measures used in several disability studies provide reasonable estimations of important subgroups of people with disabilities. We also show that the timing and structure of specific questions affects disability prevalence rates and influences observed outcomes. When we use comparable definitions across panels, we consistently find that employment rates of men with disabilities have fallen from 1990 to 1996 and employment rates of women with disabilities have remained flat. The consistency of these findings across a variety of measures illustrates an important and disturbing trend of downward employment rates for people with disabilities. These findings are particularly disturbing because they suggest that the gap in employment rates between those with and without disabilities is growing.