Six and 12-month stability of teacher ratings of temperament was studied in four samples. For the two samples retested after a 6-month interval, the same teacher provided the original and retest ratings. For the two samples retested after a 12-month interval, different teachers provided the retest ratings than provided the original ratings. Four indices of stability were investigated for each sample: cross-rank stability, within-person stability , absolute score stability, and factorial stability. Factorial stability was demonstrated for all samples. For the other three indices of stability, 6-month stability was moderate to high, and significantly higher than the 12-month stability. The general pattern of results is comparable to temperament rating data from parents, with specific coefficients being somewhat higher.