This paper presents two new methods for analysis of a technical state of large-panel residential buildings. The first method is based on elements extracted from the classical methods and on data about repairs and modernization collected from building documentations. This is a new unique method. It is easy to use, does not need expertise. The required data could be extracted easily from building documentations. For better accuracy the data from building inspections should be applied (in Poland inspections are made every year). The second method is based on extracted data processing by means of the artificial neural networks. This method could have lower accuracy but it is less prone to data errors.