This article reports two primary school teachers' perceptions of the effectiveness of lessons based on a problem-solving intervention. The intervention included enabling and extending prompts, independent student struggle time initially and time to share problem-solving strategies at the end. The intervention had two versions: one included whole class prompts and teachers anticipated students' responses before teaching; the other without these features. Each teacher implemented two lessons in year 1/2 composite classes, with one lesson common. Teachers identified positive impacts of the intervention including providing support for students, extending students' thinking and providing positive challenge during problem-solving. Struggle time was believed to negatively impact some students' resilience and confidence; both teachers deviated from the intervention to reduce struggle time. Students used more problem-solving strategies when struggle time was included compared to when the teacher modelled an approach for solving. There was a tension for teachers between providing time for students to struggle and preserving some students' confidence. One teacher facilitated student share time in the middle of one lesson, allowing students to experience both struggle and success; this compromise could address the tension. Overall, the intervention was perceived to positively impact teaching practice.