Recently, bicycle volumes have been noticeably increasing in Korea. However, relatively few studies regarding bicyclist safety have been published. Since understanding the factors which strongly influence bicycle accidents at urban signalized intersections is very valuable in improving bicyclist safety, this study developed a regression model to identify factors associated with bicycle crashes using data from 112 signalized urban intersections. We found that the Poisson regression model is suitable for identifying critical factors, which are as follows: the increase of average daily traffic volume and average daily bicycle volume tend to increase bicycle accident. When width of passenger road (side walk) is increasing, the probability of bicycle accident is decreasing. If traffic islands exist at intersection, likelihood of bicycle accident is increasing as well. We note that these factors must be considered when evaluate bicyclist safety at signalized urban intersections in Korea. Furthermore, these variables should be carefully considered for future bicyclist safety researches.