Current research on teaching mathematics to students with intellectual disability (ID) has demonstrated that, given evidence-based instruction, these students can acquire a wide range of skills. However, very little research has been conducted in inclusive general education (GE) classrooms and no studies have utilized the GE teacher as the intervention agent (Spooner, Root, Saunders, & Browder, 2018). This study utilized a multiple baseline design to evaluate the effects of GE teacher-delivered embedded instruction on two main variables: (1) students' ability to solve word problems, and (2) generalization across people and materials and to untaught word problems. The data indicate that the embedded mathematics instructional package was effective in teaching students to (1) solve additive word problems with sums less than five, and (2) to generalize to untaught word problems and across people and materials. Limitations and suggestions for future research are also discussed.