This paper addresses the security challenges of statistical databases (SDBs) in protecting confidential information from disclosure. It classifies security-control methods into four general approaches: conceptual, query restriction, data perturbation, and output perturbation. The paper identifies criteria for evaluating these methods and discusses their performance. It highlights that no single method prevents both exact and partial disclosures, but some perturbation-based methods can prevent exact disclosure while allowing statistical disclosure control. The paper recommends future research should focus on developing methods that prevent exact disclosure, provide statistical disclosure control, and avoid bias and the 0/1 query-set-size problem. The paper also discusses new types of threats and concludes with recommendations for future research and practical applications.This paper addresses the security challenges of statistical databases (SDBs) in protecting confidential information from disclosure. It classifies security-control methods into four general approaches: conceptual, query restriction, data perturbation, and output perturbation. The paper identifies criteria for evaluating these methods and discusses their performance. It highlights that no single method prevents both exact and partial disclosures, but some perturbation-based methods can prevent exact disclosure while allowing statistical disclosure control. The paper recommends future research should focus on developing methods that prevent exact disclosure, provide statistical disclosure control, and avoid bias and the 0/1 query-set-size problem. The paper also discusses new types of threats and concludes with recommendations for future research and practical applications.