OBJECTIVE： To provide reference for the study of optimal sampling points in clinical pharmacokinetics. METHODS： The literatures about optimal sampling points in clinical pharmacokinetics were searched from CNKI， Wanfang database， VIP， PubMed， Medline， ScienceDirect and other databases during Jan. 2011-Jun. 2016 using “Bayesian estimate（s）”“Bayesian estimator（s）”“Bayesian analysis”“Limited”“Optimal”“Sparse”“Minimal sampling” as retrieval words. The systematic analysis and evaluation were conducted. RESULTS： A total of 1 Chinese literature and 13 English literatures were involved respectively. The drugs they focused on were mainly immunosuppressive agents，antiviral drugs， antibiotics， pediatric individualized medication， etc. Multiple linear regression （MLR） was still the most widely used method in China，while maximum a posteriori Bayesian （MAPB） method was more popular in foreign studies. MLR equation was simple and easy to use， but the sampling was very strict. MAPB method could be completed with less sampling points and sampling time； it was more suitable for clinical practice， but needed professional software. The precision and accuracy of the two methods were similar. The research methods of optimal sampling strategy were quite different but all included 4 steps as prior information ganining， reference value determination， sampling point optimization， prediction capability verification. CONCLUSIONS： MAPB method requires less sampling points and it results are relatively accurate and reliable. It is more suitable for clinical practice and optimal sampling study of clinical pharmacokinetics.