Synthetic small interfering RNAs (siRNAs) open promising new therapeutic perspectives in acute and chronic pathologies. A number of experiments in mice demonstrated the ability of naked siRNAs injected under a normal pressure to trigger gene silencing in vivo, translating into a measurable phenotype. We focus in this review on the information that we can gain from these experiments, and discuss how the specificity of the gene silencing in vivo can be controlled. Because the activity of most drugs increases with the dosing, we are prone to consider that increasing the concentration of siRNAs within cells enhances the efficiency and the duration of the silencing. However, because RNAi is a saturable process, and because increasing the siRNA concentration into cells can induce undesirable side effects, this must be demonstrated. We compare in this review the methods used to quantify and study the biodistribution of siRNAs in living animals, and discuss how these methods can help in designing for each model and each siRNA the most adequate protocol to silence a cognate target gene in vivo.