Focusing on the use of category systems in classroom observation, this report summarizes factors that may significantly affect the validity of cross-national and cross-cultural comparisons of classroom behaviors. Category systems measure well-defined behaviors by recording events observed at specific intervals or as they begin and end. Areas of concern related to the use of this strategy in multicultural comparisons include cost, identification of comparable populations, resistance by some teachers to observation, observer effects on classroom behaviors, the selection of the specific behaviors chosen for observation, the classification of significantly different events in the same category, limitations in the number of behaviors measured, and theoretical and cultural biases. The encoding and decoding processes for category systems also raise several concerns, including inter-investigation reliability, observer accuracy, subjectivity, distortions in recording procedures, problems in context and meaning, and failure of personnel to follow project protocol. Those designing future studies should consider the fact that the effects of many of these factors can be controlled or eliminated. Moreover, persons intimately familiar with the classrooms of the cultures to be compared should become an integral part of almost any investigation in which category systems are used. (Author/WD)