This paper presents a quantitative meta-analysis of studies evaluating, by means of the Bayesian method, the safety effectiveness of different bicycle facilities at road junctions. This is preceded by a discussion of background theories and an up-to-date presentation of today's knowledge of the cyclist's safety at intersections; alternative layouts in Scandinavia, and Sweden in particular, are shown. The project consists of a literature survey as well as interviews with experts and cyclists, and an attempt to weave together these different sources in estimating the effect of a particular layout. In summary, few studies from the Scandinavian countries exist that have treated this area with an acceptable methodology. Combined results, with the Bayesian technique, are therefore presented for only one layout comparison: accident risks for cyclists at signalized intersections with and without a cycle path. The results of this aggregation may be unreliable as well, due to deficiencies in the studies. New field studies should be initiated. The "experts'" prior opinion was that the introduction of the cycle path would, on average, increase the risk by about 20%, while interviewed cyclists considered that a cycle path would decrease the risk by about 20%. The conclusion that can be drawn so far from combining results shows that the most likely effect of introducing a cycle path is that the risk will increase by about 40% for a passing cyclist. The probability that the effect will be the opposite, i.e. that the risk will decrease, is very small (about 2%). These combined results are based primarily on cross-section studies, where the layouts have not been "randomly allotted". It is therefore likely that the risk increase may have been overestimated. The probability that the "real" effect is accident reduction is therefore somewhat greater than 2%. How much greater cannot be estimated from these studies.