The officially reported impressive rises in standards in mathematics and English in primary schools since 1995 are challenged. The article looks at the increases from four different perspectives. First, the general pattern of change is considered. Secondly, the statutory test data are compared with the results from several different studies. These indicate a complex pattern with clear rises in standards, but not as strong as the official data suggest up to 2000. Since 2000 the official data have shown little change and this is largely confirmed by independent tests. Thirdly, the standard setting procedures are considered and mechanisms by which the observed patterns could have been produced are set out. Finally, parallels are drawn with the experience in Texas where an apparently remarkable set of data was shown to be largely illusory. A case is made for an independent body to be set up with the express and sole purpose of monitoring standards over time.