The goal of this study was to clarify the causes of natural stone pavement failures, and to assess which failure types are critical from the points of view of traffic safety, maintenance and aesthetics. There were three methods of study: The first method was a literature research on building and repairing instructions for natural stone structures. The second method was a field survey of the current state of the natural stone streets in Helsinki and their damages. The third method was interviewing experts on the probable causes of damages to natural stone streets and on how the development of these damages could be slowed down and prevented. The results of the field investigations were analyzed by comparing the failures of inside structures. According to the literature research and interviews the reason for the natural stone pavement structure failures is often the settling of the sandy road bed, which is caused by the huge traffic load of the street. The city of Helsinki has no specific instructions on designing and building natural stone structures on heavily trafficked streets, albeit having four instruction manuals on natural stone street design. That is why the need to specify the design instructions is obvious, and Helsinki could use the natural stone street design instructions of Scotland, for example, as the basis for its own instructions. It was revealed in the interviews that, especially in Helsinki, failures are the consequence of defective supervision of excavation works, and poor work quality is accepted too easily. That is why supervision should be improved. Furthermore, according to the interviews a missing stone is the most critical failure type and should be accounted for at once. Missing stones create sharp impacts against cars and also further the decrease of bearing capacity. Longitudinal and lateral deformations and individual sags in large quantities indicate the need of renovation of natural stone streets. The results from the field investigations were compared using statistical analysis between failures of different street sections. The differences between cube stones and Belgium stones were found to be statistically insignificant. However, opinions in the interviews were that cube stones, due to their smaller dimensions, are weaker than Belgium stones, which is why using them should be avoided. Also, an important finding statistically was that streets with gradients higher than 8 % were found to be more damaged than gentler streets with gradients lower than 8 %. Therefore paving steep streets with stones should be avoided.