A large body of research has shown that for novice learners, instruction that relies more heavily on worked examples than on problem solving, is more effective for learning as shown by higher test performance. Moreover, this beneficial effect is often obtained with less acquisition time and lower cognitive load during acquisition and test phase. However, most of this research has been conducted in laboratory settings with college or university students and a control condition consisting of problem solving without any additional support. The present study, using a quasi-experimental design, investigated the effects of implementing worked examples in an existing primary school mathematics curriculum in which a realistic mathematics teaching method is used, during a 3-week period. The results showed no significant differences in test performance or cognitive load; however, the worked examples group attained this level of performance with significantly less acquisition time. (Contains 2 tables.)