Background: The prediction models for infertility treatment success have presented since 25 years ago. There are scientific principles for designing and applying the prediction models that is also used to predict the success rate of infertility treatment. The purpose of this study is to provide basic principles for designing the model to predic infertility treatment success.

Materials and Methods: In this paper, the principles for developing predictive models are explained and then the design of such models in infertility treatments is described in more details by explaining one sample.

Results: The important principles for models that firstly are described are: identifying and defining the purpose, expected function of model, input data that will be used to develop a model: type of intervention or diagnostic procedures that can lead to changes in the samples and output definition or expected result of model function. Further, characteristics of predictive factors in final model, drawing the information flowchart, internal and external validation and attention to the analysis programme of results are the important subjects that have been described.

Conclusion: If predictive models are used properly, can help treatment team and patients to achive best treatment in ART.