Fires in buildings are dangerous in many ways, one big hazard is the risk of building collapse. This is a hazard both to occupants in the building and to the fire service, who might have to enter the building for their firefighting. The consequences of a building collapse can become severe, both in terms of damage to the building and in terms of injuries to people. Despite this, there are few methods to determine how stable a construction is, when exposed to stress from a fire. This thesis has tried to find connections between occurred events where buildings have collapsed, both by a literature study and by looking into some occurred collapses. This has been done by analysing some specific events more deeply, to find out what went wrong in that specific case, and by looking into a broader base of occurred building collapses which has been presented as statistics. Finally contact with personnel from two different fire brigades was taken, to perform a questionnaire and collect information of how they work with the dangers in reality. All of these studies has shown that there are some connections between which buildings collapse more often than others, one example is that small houses collapse more frequently than residential buildings and that roofs is the part of the building which collapses most often. This is confirmed by some of the people from the fire brigades, who thinks it reflects their own experiences regarding building fires. Even though some connections like these can be shown, they can have several explanations as for example, there are many more small houses than residential buildings in Sweden which makes it natural to believe that they collapse more often. Still, some connections are clear and this information can be used to base further studies on and perhaps gain more knowledge to be able to make better judgements on whether a construction is stable or not.