Forty-seven elementary school teachers and principals participated in a six-week summer workshop to prepare them for teaching American Association for the Advancement of Science (AAAS)-Science A Process Approach. The workshop participants had a pre-test mean score of 66.7 on Form A of the AAAS Science Process Measure for Teachers. The post-test mean was 80.0. Twenty-six of the participants taught Science-A Process Approach in K-3 during the following year. Four sets of competency tasks, taken from AAAS Competency Measures, were used in determining gains in student achievement over a period of a year. The research design provided for the pretesting in September of 15 control students and 120 experimental students randomly selected at each grade level, K-3. Post-tests were administered to both groups in May, 1968. The Mann-Whitney U statistic was used as a test for differences between the two groups. The findings were: at all four grade levels there were no initial significant differences at the .05 level between the control and experimental groups. Post-tests in all four grades indicated that significant achievement differences existed in favor of the experimental group. (BR)