Using Program for International Student Achievement (PISA) 2009 data we study the transfer of knowledge among reading, mathematics, and science among Turkish students. Both Science and Reading are significant predictors of Mathematics scores, although clearly Science is a much stronger predictor; the transfer from Science to Mathematics is much greater than is the transfer from Reading to Mathematics. SCHOOLID is the single strongest predictor of Mathematics outcomes, likely reflecting the importance of socioeconomic and regional or urban/rural differences in the quality of education available to students. Both Mathematics and Reading are significant predictors of Science scores, although Mathematics is a stronger predictor; the transfer from Mathematics to Science is greater than is the transfer from Reading to Science. SCHOOLID is a weaker predictor of Science outcomes than are Mathematics scores, suggesting that the importance of socioeconomic and regional or urban/rural differences in the quality of education available to students may have slightly less consequence for Science outcomes than does the transfer effect from Mathematics to Science. Both Science and Mathematics are significant predictors of Reading scores, but the transfer from Science to Reading is much more robust than the transfer from Mathematics to Reading. SCHOOLID and Science are nearly identically strong predictors of Reading outcomes, suggesting that the importance of socioeconomic and regional or urban/rural differences in the quality of education available is on a par with the Science transfer to Reading. Implications of these findings are discussed.