摘要: Using the evaluation database on world collection of rice, Oryza sativa, conserved at International Rice Research Institute, different sampling strategies for choosing germplasm were compared. Random, stratified, sequential and analysed sets chosen frequency finding resistance to rice pests, brown planthopper, green leafhopper whitebacked diseases, bacterial blight blast The geographically restricted javanica race was also compared in sets. results indicate that where no prior information is available choose evaluation, same sample number, representing broad genetic diversity are preferable other strategies.