The hypothesis of this project is that the safety of cyclists can be improved by increasing their visibility in traffic. This is explored by testing whether or not a high-visibility bicycle jacket (colour and reflectors) will increase the safety of cyclists. The project has been carried out as a randomized controlled trial with 6,800 volunteer cyclists. After random selection, half of the group – the test group – got the bicycle jacket at once and promised to wear it each time they biked during a year. The other half of the group composed a control group that got the bicycle jacket after the closing of the project, i.e. after a year. The safety effect of the bicycle jacket was analysed by comparing the number of self-reported accident for the test and control group. The self-reported accidents showed that the test group had 38 % fewer personal injury accidents with other road users – so-called multi-party accidents – than the group who did not wear the bicycle jack-et. If one only looks at accidents between participants and vehicles, the difference is 48 %. The differences are statistically significant at 5 % level. In the test group, 37 % of the involved parties in an accident report-ed that they were not wearing the bicycle jacket or any other bright-coloured garment when the accident occurred.