Prevention of pressure sores is a central task of nursing. For ethical and economic reasons correct assessment of the risk of pressure sores is a prerequisite of adequate prevention of pressure sores. In Germany the Norton scale is one of the most widespread tools for assessing the risk of pressure sores. For practical purposes the original translated version of the Norton scale is in use, as well as an extended modified version. Validity and reliability of these two versions are in the early stages of examination in Germany. The present descriptive longitudinal study aims at determining the predictive value of both versions of the Norton scale in nursing homes. In a convenience sample of 146 individuals from five nursing homes residents from the age of 60 were assessed by professional nursing staff with the aid of both versions of the Norton scale at weekly intervals. Simultaneously the skin status was recorded on both sampling days. The results exhibit high values of sensitivity (91%-95%) and low specificity (6.2%-20.6%) in both scales. Both scales tend to overrate the risk of pressure sores. The values indicate that the original Norton scale assesses the residents free of pressure sores better than its extended version. Because of the convenience-sample the results are not representative and apply only to the investigated setting and population. The predictive validity of the original and modified version of the Norton scale is insufficiently verified, so there is great need of further investigation of the validity of the two versions. Application of assessment scales in nursing should be supplemented by professional expertise of the nursing staff to counterbalance assessment errors of the tools.