Pre-election opinion polls in Tanzania are relatively new. They came with the establishment of the multiparty system in 1992. So far, the known pollsters which involved in the previous elections include the Tanzania Citizens’ Information Bureau (TCIB), SYNOVATE, and the Research and Education for Democracy in Tanzania (REDET). To these, there is a proliferation of electronic as well as “sangoma” polls. Yet, the polling industry has remained incredible. In every previous election, pollsters were criticised by key stakeholders as being biased. Indeed the 2010 elections marked the climax. Despite such criticisms, no one went a step further to examine the scientific side of the polling industry. This work breaks the ground by providing provocative insights on sampling procedures, framing of questions, and reporting by pollsters. It notes that the methodology by individual pollsters was either by default or design compromised leading to controversial polling outcome. The book is a wakeup call to pollsters to rethink the most acceptable standards. It may be useful to pollsters, political parties, researchers, and college students.