The study included field data collected from outbreak areas of Norway spruce beetle Ips typographus L., which were used to identify stand conditions associated with outbreak populations. In 2001-2002 data from 100 infested and 100 uninfested plots were collected from eight Forest Districts of State Forests and three National Parks in northeastern Poland. Among 17 analyzed variables only the number of trees for all species resulted in a significant difference between infested and uninfested stands. To develop probability of infestation models, a statistical technique called Classification and Regression Trees (CART) was used. Results of classification models indicate increased probability of infestation when stand density index for all species of trees present declines below 14.1 per plot (0.02 ha) (about 0.6 of common density index used in polish forestry) and for stands with a high percentage of spruce (> 92.2 percent).