A stratified random sample of volunteer participants (N = 75) aged 9 to 14 was drawn from 16 public school districts' gifted programs, including as many females (n = 28) and children from groups traditionally underrepresented in gifted programs (n = 18) as available. Participants were randomly divided into an experimental (n = 38) and a control group (n = 37) for an intervention study. All participants took the "Cognitive Abilities Test (Form 6) Verbal Battery" and the "Project TALENT Spatial Ability Assessments". The experimental group participated in a simulation of the FIRST LEGO League (FLL) competition for 20 hr total over five consecutive days. All participants took the spatial measure again. Experimental males evidenced significant and meaningful gains in measured spatial ability (Cohen's d = 0.87). Females did not evidence significant gains in measured spatial ability. This may be due to sampling error, gender differences in prior experience with LEGO, or differences in facets of spatial ability in the treatment or measurements. (Contains 5 tables and 1 figure.)