A two-stage randomized response (RR) technique was proposed by Chang and Huang (2001) for estimating \(\pi\) , the proportion of individuals in a population that possess a certain sensitive attribute \(A\) and also estimating \(T\) , the sensitivity level of a question, which is defined as the proportion of individuals who possess the sensitive attribute \(A\) , answer “yes” to the direct sensitive question “Do you possess the attribute \(A\) ?”. Huang (2004), Marcheselli and Barabesi (2006) extended the method further by applying different RR methods in two stages and found that the two-stage RR procedure is more efficient than the single-stage RR procedure for estimating \(\pi\) . Quatember (2012) extended the two-stage procedure to the general \(h( \ge 2)\) stages. Arnab (2023a) proposed two-stage RR procedures for complex survey designs and showed that the proposed procedure performed better than the corresponding single-stage RR procedures. He also showed that the results derived by Huang (2004), Barabesi and Marcheselli (2006) for the two-stage RR procedures can be obtained from his method as special cases. In this chapter, the works of Chang and Huang (2001), Huang (2004), Kim and Warde (2005), Marcheselli and Barabesi (2006), Quatember (2012), and Arnab (2023a) are reviewed.