In the current study we sought to determine if children improved their self-reported personal/social skills across an academic year. Elementary students (N=720) in physical education classes were taught by two groups of teachers (N=30) who were assigned to different training protocols in a yearlong curricular inservice program aimed at integrating the Exemplary Physical Education Curriculum (EPEC) into an urban district’s curriculum. Training group A received retraining in the EPEC curriculum. Training group B received the same retraining as well as inservice training throughout the year. Boys (n=350) and girls (n=369) in first (n=361) and third (n=359) grades completed personal/social development portfolios in the fall and spring. Most of the children (80%) had an African-American heritage. Portfolios consisted of 5 questions upon which an overall personal/social development behavior mean was calculated. An example of a question is “I put the balls in the basket after the teacher tells me to”. Students responded using the response options of “I do that” “I sometimes do that” and “I never do that.” The portfolios have previously shown that they can produce reliable and valid scores in children. Teachers also rated the students on their social development using a Likert-like scale ranging from 1-5. Repeated measures ANOVA was used to determine if significant changes occurred over time and by teacher training group, grade, and gender. Children’s personal/social behavior scores were significantly higher at the end of the year [F(1, 718) = 93.21, p < .01, partial eta squared (çp 2) = .12]. Students from the two teacher training groups were significantly different [F(1, 718) = 4.13, p < .01, çp 2 = .006]. Students from teacher training group A had lower mean scores (better behaviors) at pretest and posttest. First and third grade students were also significantly different [F(1, 718) = 23.01, p < .01, çp 2 = .031], with first grade students reporting better personal/social behaviors at pretest and posttest than third grade students. Finally, boys and girls were also significantly different [F(1, 718) = 15.44, p < .01, çp 2 = .021], with girls reporting better personal/social behaviors at pretest and posttest than boys. Teachers’ ratings and students’ mean personal/social behavior scores were significantly correlated. These findings suggest that all of the students improved in their personal/social behaviors over the year. Younger students (i.e., grade 1) and girls reported better behaviors than older students (i.e., grade 3) and boys.Keyword(s): elementary education, professional development, research