When a log is missing in a drilling hole, geologists hope to deduce itfrom others logs available in another part of the hole or in aneighbouring hole, in order to define the lithologic facies of the hole.This paper presents a neural network method to predict the missing log'smeasure from the other available log's measures. This method, based onMulti-Layer Perceptron (MLP) acts as a non linear regression method forthe prediction task and as a probability density distributionapproximation for the outlier rejection task. The result obtained whenapplied to actual log's data for prediction and rejection are presentedin a separate section. The last section is dedicated to a non supervisedneural method in order to reconstruct the lithologic facies of theconcerned hole. This last experiment allows to validate and interpretthe different results of the proposed methods.