School psychologists rely heavily upon teachers' evaluation of children's classroom behavior as part of their assessment and intervention processes. However, some evidence suggests that teachers' ratings may fluctuate during the academic year, with higher levels being reported later in the year. This study assessed the comparability of different teachers' ratings of children's externalizing behavior from fall and spring intervention cohorts. Teachers of children from divorced families (N=240, ages 9 to 12) completed pretest, posttest, and six-month follow-up ratings. Two separate ANOVAs revealed that changes in mean levels were not significantly different between the fall and spring cohorts for either the pretest/posttest or pretest/follow-up comparison. Ratings obtained from children's current teachers within the first month of school were compared to those completed by the children's teachers from the previous school year. Pretest levels of externalizing behavior were higher when reported retrospectively by the previous teachers. However, neither the previous teachers' nor the current teachers' fall ratings were significantly different from the spring teachers' pretest ratings. Thus, school psychologists may elect to obtain ratings from either the previous or current teacher early in the fall of a new academic year in the assessment of children's behavior. (Contains 2 tables and 12 references.) (JDM)