Recent analyses of state-level proficiency standards under No Child Left Behind indicate that states' benchmarks for determining whether a child is proficient in reading and math are widely divergent. This article explores whether the capacity of states to employ different proficiency standards imposes different performance standards across demographic groups. Using a newly devised metric that allows for interstate comparison of state-level proficiency cut scores, along with aggregated district-level demographic data, this article provides a descriptive analysis of the distribution of proficiency standards across demographic groups to determine whether some groups systematically face higher or lower proficiency cut scores. The findings indicate that while an "expectations gap" does exist, it does not operate consistently across categories of race, ethnicity, and poverty status. The article concludes by reviewing the trade-offs required under three possible policy responses to these findings. (Contains 3 figures, 6 tables and 6 notes.)