Dr. Gary W. Phillips & John A. Dossey [2.5MB PDF Report]:

Students in six major U.S. cities are performing on par or better in mathematics than their peers in other countries in grades 4 and 8, according to a new study by the American Institutes for Research (AIR). However, students from five other major cities are not faring as well, and overall, U.S. student performance in mathematics falls off from elementary to middle school grades — and remains behind many industrialized nations, particularly Asian nations.
The AIR study offers the first comparison between students from large U.S. cities and their international peers. The study compares U.S. 4th grade students with their counterparts in 24 countries and 8th grade students with peers in 45 countries.
“Globalization is not something we can hold off or turn off…it is the economic equivalent of a force of nature… like the wind and water” (Bill Clinton)
If you are a student today competing for jobs in a global economy, the good jobs will not go to the best in your graduating class–the jobs will go to the best students in the world. Large urban cities are intimately connected to the nations of the world. Large corporations locate their businesses in U.S. cities; foreign students attend U.S. schools; and U.S. businesses export goods and services to foreign nations. Large urban cities need to know how their students stack up against peers in the nations with which the U.S. does business. This is especially important for students in the fields of science, technology, engineering, and mathematics. The students in these fields will allow our future generation to remain technologically innovative and economically competitive.
This report provides a comparison of the number of mathematically Proficient students in Grades 4 and 8 in 11 large cities in the United States with their international peers.
This comparison is made possible by statistically linking the National Assessment of Educational Progress (NAEP) in 2003 and the Trends in International Mathematics and Science Study (TIMSS) in 2003 when both assessments were conducted in the United States in the same year and in the same grades.
After the statistical linking was completed, it was possible to compare the most recent NAEP results (from 2007) to the most recent TIMSS results (from 2003). How the United States compares to the overall international average.
At Grade 4, five countries (Singapore, Hong Kong SAR, Chinese Taipei, Japan, and the Flemish portion of Belgium) performed significantly better than the United States (Figure 1). However, the United States (at 39% proficiency) performed better than the international average (27% proficiency) of all 24 countries (Figure 13).
At Grade 8, eight countries (Singapore, Hong Kong SAR, Republic of Korea, Chinese Taipei, Japan, Belgium (Flemish), Netherlands, and Hungary) performed significantly better than the United States (Figure 1). However, the United States (at 31% proficiency) performed better than the international average (21% proficiency) of all 44 countries (Figure 14).
….
Because of the persistent requests of urban school districts, the U .S . Congress authorized NAEP to assess, on a trial basis, six large urban school districts beginning in 2002 . Since then, more districts have been added, resulting in 11 school districts in 2007 (and plans are underway to include even more districts in the future) . The urban school chiefs in these 11 large school districts, which voluntarily participated in the 2007 NAEP, recognized the global nature of educational expectations and the importance of having reliable external data against which to judge the performance of their students and to hold themselves accountable . They should be commended for their visionary goal of trying to benchmark their local performance against tough national standards. National standards provide a broad context and an external compass with which to steer educational policy to benefit local systems . The purpose of this report is to further help those systems navigate by providing international benchmarks.

Clusty Search: Gary W. Phillips and John A. Dossey.
Greg Toppo has more:

Even if the findings are less-than-stellar, he says, they should help local officials focus on improving results.
“In that sense, I think it could be a very positive thing to use in-house, in the district, to keep their nose to the grindstone,” says Kepner, a former middle- and high-school math teacher in Iowa and Wisconsin.”If they can show they’re improving, they might be able to attract more companies to a system that’s on the move.”
Phillips says the findings prove that in other countries “it is possible to do well and learn considerably under a lot of varied circumstances — in other words, being low-income is not really an excuse when you look around the rest of the world.”

Math Forum audio & video.