This chapter highlights the key assumptions underlying Randomized Control Trials (RCTs) and illustrates them with regard to the practice of RCTs in the realm of child and adolescent development. Given the prominence of RCTs in policy research, we analyze the possible ways in which these assumptions might not be met by single randomized experiments, thereby making it difficult to draw valid causal inference from single studies. We frame this discussion within the categories of internal validity, statistical conclusion validity, construct validity, and external validity and address the debate about the "gold standard" status accorded to RCTs.