We evaluated the cost-effectiveness of Struggling Schools, a user-generated approach to Comprehensive School Reform implemented in 100 low achieving schools serving disadvantaged students in a Canadian province. The results show that while Struggling Schools had a statistically significant positive effect on Grade 3 Reading achievement, d = 0.48 in 2005-06 and 0.60 in 2006-07, the program was not cost-effective when compared to two alternatives: (1) The cost of bringing one student to the provincial achievement standard was more than 25% higher in Struggling Schools than in the status quo; and (2) The cost-effectiveness ratio (i.e., effect size per $1,000 of incremental cost) was lower in Struggling Schools than in Success For All. Struggling Schools would have been deemed to be cost-effective if different choices had been made, especially in (a) the calculation of costs (e.g., the inclusion of donated time), (b) the decision rules for declaring cost-effectiveness, and (c) the studies used to access comparative data.