The research described in this document investigated the feasibility of matching teachers with their preferred style of teaching. Ninety teachers working with sixth-grade students were randomly assigned to either a highly structured or a low-structured curriculum. Teaching effectiveness (judged by student achievement) was compared in three situations: training in the preferred style, training in the non-preferred style, and teaching in the preferred style without training. The main effects were analyzed with t tests on the outcome measure. Training effects were analyzed by competency comparisons across treatments, using means and standard deviations of observer ratings. Results indicated that teaching assignments and inservice training can be adjusted to teaching style. Low structured curricula may be destructive to children and teachers with high anxiety levels. The research reported in this document was subject to a major technical difficulty and as a consequence the speculations should be qualified. Despite this technical weakness, the report is being circulated because the work undertaken may be of interest. (Author/MBM)