The results presented in this paper are part of a current comparative study of genetic and environmental influences in three educational settings: Stockholm, Jerusalem and the Israeli kibbutz. We specifically wanted to investigate whether a more restrictive educational setting would decrease genetic influences. Here we report on comparisons of cognitive performance measures at several time points for twins/controls, boys/girls and within-pair similarity in MZ, DZ and controls. The tests used were the Raven Progressive Matrices, verbal, reading comprehension and arithmetic. The results show no differences between twins and controls, whilst gender differences seem to be smallest in the Stockholm sample and largest in Jerusalem. A pattern of genetic influences on cognitive performance was also clearly visible in Jerusalem. In Stockholm shared environmental influences at home and at school seemed even stronger than in the kibbutz. No consistent differences were found between tests or occasions.