To investigate differences in students' writing development due to teacher modeling, a study examined students enrolled in whole language classrooms in grades 1 and 2 in two separate school districts. Students enrolled in independent classrooms where teachers did not provide writing models were in a school in a lower middle class neighborhood of Vancouver, British Columbia. Approximately 65% of the students were from families in which English was the second language (ESL). Modeling classrooms were in a school in a middle class neighborhood of Richmond, British Columbia, a suburb of Vancouver. Approximately 30% of the students were ESL. Teachers in independent classrooms asked students to write on the first day of class. Students read their writing to the teachers who responded positively. Teachers also kept notes about students' writing. As students began producing more mature texts, teachers kept a chronological log of their conferences. Teachers in the modeling classrooms asked students what they would like to write, then wrote the sentence(s). Often students emulated the teacher's model. In addition, teachers wrote comments containing the essence of what students indicated they had written. For both types of classes, students' writing for first and second grades was photo-copied. Each writing product was analyzed for number of words, number of t-units (independent clauses with all attendant modifiers), quality, and topic choice. Quality was measured on a nine-point experimental scale developed to reflect writing structure and quality. Findings revealed that both strategies produced eager, competent writers. (Two figures are included.) (MM)