In the spring of 1973, a standardized test of reading and mathematics skills was administered to 564 grade 3 pupils distributed among two open plan schools (155 pupils) and seven conventionally constructed schools (409 pupils). The groups were matched on IQ, geographic location, the academic qualifications of their teachers, and average class load. This was the third and final year of a longitudinal ability and achievement study of this pupil cohort. As in the previous two years, the marginal differences in achievement as disclosed by test results were neither consistent nor significant. Over the three schools years, beginning in September 1970, no significant differences were found in the academic ability and scholastic achievement levels between the pupils in the two sorts of architectural settings. No consistent patterns were discernible; some differences might be attributable to factors other than architecture, since it was impossible to control all the variables (e.g., years of teaching experience) that might reasonably have affected the learning situations under study. (Author/MLF)