Abstract: The application of two multifactor experimental designs,orthogonal design (OD) and uniform design (UD),for developing pesticide microemulsion (ME) were investigated through a case study of abamectin 20 ME.In experimental design,the total dosage of surfactants was fixed,each of the used surfactant was as independent variable except one,and co-surfactant and antifreeze were also as independent variables.Transparency temperature section (TTS) of ME was tested as response in OD and UD experiments,and its low point temperature (TL),high point temperature (TH) and range (TR=TH-TL) were selected as dependent variables.Analysis of variance and rang R were introduced to processing OD experiment data,and TH was selected as dependent variable; TTS of abamectin 20 ME optimized by OD was -15.0℃~66.5℃.Partial Least Squares Regression was introduced to processing UD experiment data,and TH and TR were selected as dependent variables and optimized together during modeling; two regression equation models were obtained,and R2 of models were 0.9945,0.9851,respectively; the prediction values of TH and TR were 71.5℃ and 87.1℃,respectively,while actual tested values were 71.6℃ and 87.3℃,respectively.Turbiscan scanning showed that the preparations optimized by OD and UD were very stable during stored 14 d successively at 54℃.Results showed that: (1) OD and UD could be successfully applied to pesticide ME development.The two experimental designs could substantially lighten formulators' workload,and help formulators to quickly obtain recipe with good performance.(2) TTS was a appropriate response variable for OD and UD,and it could be split into three dependent variables: TL,TH and TR.