This paper is concerned with the use of Bayesian analysis in women health diagnosis. Illustration example is given on real data collected from women at high risk pregnancy. A suggested coding for the classification in a multidimensional symptom space for diagnostic aid of congenital pregnancy diseases is given. This method enables us to classify factors that effect on pregnancy and the result is the women pregnancy in the state of high risk pregnancy, and the side effect of this factors on women pregnancy and fetal can be computed by use of Bayesian analysis. Probability, ratio, average and conduction for every diagnosis were computed. Risk factors, loss function and posterior expected loss were defined and computed.