Embedded multimedia refers to teaching methods that embed video content within teachers' lessons. The research of Mayer and others has suggested that multimedia instruction can enhance learning by using the capacity of both visual and verbal memory systems. The present study is an evaluation of embedded multimedia in a year-long randomized clinical trial comparing first graders who learned beginning reading using the Success for All program either with or without embedded, brief video components. A study involving 394 first graders in 10 high-poverty schools found significant positive effects on the Woodcock Word Attack scale, controlling for pretests, in HLM analyses with school as the unit of analysis. The results provide partial support for the utility of embedded multimedia as a component of beginning reading instruction.