Research into the stability of school effects has revealed a substantial amount of instability across grades, but the hypothesis that this instability can be attributed to teacher effects has not yet been tested. In the present study it has been investigated to what extent teacher changes can account for the instability of student achievement across grades. For that purpose a dataset was analyzed which contains information about students who were followed throughout their last two years in elementary education. One group of students had been taught by one and the same teacher during these two years, whereas the other group had been taught by two different teachers. The analyses did not focus exclusively on student achievement. The (in)stability of teacher perceptions and teacher goals was investigated as well. All in all, the (in)stability across grades was analyzed for nine variables. The instability across grades was found to be significantly smaller for the students with the same teacher for four of the nine variables, but the differences between both groups were not very large. It should be noted, though, that the analyses relate only to the (in)stability at the individual level. The effect of teacher changes at the (in)stability of classroom averages may be quite different.