This formative study investigated three resource/classroom teacher collaborations which focused on providing classroom instruction to low-achieving grade one readers. Difficulties encountered when instruction is provided in a separate resource setting include: (1) Students may be stigmatized socially; (2) Learning often does not transfer back into the classroom; and (3) Classroom teachers may not be able to reinforce learning because they are unaware of resource room activities. Few students receive benefits when resource teacher assistance is restricted to a 'pull-out' delivery system. In a urban elementary school, the researcher/resource teacher collaborated with grade one classroom teachers to assist low-achieving readers chosen through the results of whole class screening tests. Clay's Observation Survey (1993), field notes, written reflections, work samples, and running records were used to document student progress. Audio-recordings of meetings with classroom teachers and a final teacher interview were used to follow the progress of the collaboration. Findings suggested that moving resource assistance into the classroom and having the resource and classroom teachers collaborate eliminated the transfer of learning, stigmatization, and communication problems inherent in pull-out service delivery. The substantive differences among the three collaborations were based on the relationships that developed between the collaborating pairs. All of the focus students made gains in reading development, but no association was found between the quality of collaboration and individual student success. Based on a literature review, a set of characteristics central to collaboration was developed: planning, assisting, sharing expertise, trusting, and their negations. Frequency counts of these characteristics, placed in ratios, sort collaborations into a 'quality of collaboration' continuum. This approach to data analysis could be generalized for use in future studies on collaboration.