Nursing narratives are an important part of patient documentation, but the possibilities to utilize them in the direct care process are limited due to the lack of proper tools. One solution to facilitate the utilization of narrative data could be to classify them according to their content. In this paper, we addressed two issues related to designing an automated classifier: domain experts' agreement on the content of the classes into which the data are to be classified, and the ability of the machine-learning algorithm to perform the classification on an acceptable level. The data we used were a set of Finnish intensive care nursing narratives. By using Cohen's κ, we assessed the agreement of three nurses on the content of the classes Breathing, Blood Circulation and Pain, and by using the area under ROC curve (AUC), we measured the ability of the Least Squares Support Vector Machine (LS-SVM) algorithm to learn the classification patterns of the nurses. On average, the values of κ were around 0.8. The agreement was highest in the class Blood Circulation, and lowest in the class Breathing. The LS-SVM algorithm was able to learn the classification patterns of the three nurses on an acceptable level; the values of AUC were generally around 0.85. Our results indicate that one way to develop electronic patient records could be tools that handle the free text in nursing documentation.