Description: Academic achievement is moderately (r = 0.5) related to Intelligence Quotient (IQ) of students if the calculations are made on the individual level (Jensen, 1998, 277). The correlation is higher when aggregated data are used in calculations. Richard Lynn and Tatu Vanhanen (2002, 69) report correlation 0.48 - 0.84 between IQ and TIMSS (Trends in International Mathematics and Science Study) tests carried out in the eighties and the nineties. It is interesting to find out the correlation between the TIMSS 2003 results and IQ and investigate the other correlates of TIMSS results when the differences in IQ are considered. Methodology: The sources used in the research were TIMSS 2003 results (Mullis et al., 2004; Martin et al., 2004) and the data about nations IQ in the book by Richard Linn and Tatu Vanhanen (2002). Subjects. TIMSS 2003 was carried out in 46 countries with the 13-year-old students and in 25 countries with the 9-year-old students. The students tested were representative to the whole population of the students of this age in these countries. Three to ten thousand students participated in both age groups in every country. IQ tests were carried out with the samples from 129 persons to 64 000 persons in a country. Instruments. TIMSS 2003 tests measured factual knowledge, solving routine problems, and reasoning in the different areas of physics. The IQ of people in the countries was measured by Cattel, WISC-R, CPM and the other tests of mental ability. Because the IQ tests were carried out in different years, the Flynn effect was considered in calculations. Procedure. There were 42 countries for which IQ and TIMSS data were available for the eighth grade students and 23 countries with the data for the fourth grade students. The coefficients of Pearson correlation between TIMSS 2003 results and IQ were calculated. In addition, partial correlation coefficients were calculated and regression analysis carried out to find the characteristics of schooling, teacher training and economics, which might have effect on TIMSS results, if the effect of IQ is considered. Conclusions: The correlation coefficients between TIMSS 2003 test results and nations IQ is about 0.9 (Table). IQ is a very good predictor of TIMSS results and vice versa: the quality of education predicts average level of IQ. Correlation coefficients between TIMSS 2003 results and IQ Subject Grade Sub-tests Summarytest result I II III IV V Mathematics Fourth grade 0.86 0.87 0.90 0.90 0.92 0.91 Science Fourth grade 0.87 0.88 0.86 - - 0.88 Mathematics Eighth grade 0.89 0.90 0.93 0.89 0.93 0.92 Science Eighth grade 0.91 0.88 0.91 0.91 0.88 0.91 Partial correlation coefficients and regression analysis revealed no other effector of the TIMSS results in three cases of the four in the table. The TIMSS science test results in fourth grade were effected by class size and school climate besides the IQ. IQ has high correlation with economic welfare of nations (Lynn, Vanhanen, 2002). Therefore, it is important to pay more attention to the development of IQ in schools. Schooling may have a positive effect on IQ (Blair et al. 2005; Whalley et al. 2005). The active teaching methods, development of thinking skills, positive school climate, etc are important.