Although many researchers have estimated crash modification factors (CMFs) for specific treatments (or countermeasures), there is a lack of studies that explored the heterogeneous effects of roadway characteristics on crash frequency among treated sites. Generally, the CMF estimated by before–after studies represents overall safety effects of the treatment in a fixed value. However, as each treated site has different roadway characteristics, there is a need to assess the variation of CMFs among the treated sites with different roadway characteristics through crash modification functions (CMFunctions). The main objective of this research is to determine relationships between the safety effects of adding a bike lane and the roadway characteristics through (1) evaluation of CMFs for adding a bike lane using observational before–after with empirical Bayes (EB) and cross-sectional methods, and (2) development of simple and full CMFunctions which are describe the CMF in a function of roadway characteristics of the sites. Data was collected for urban arterials in Florida, and the Florida-specific full SPFs were developed. Moreover, socio-economic parameters were collected and included in CMFunctions and SPFs (1) to capture the effects of the variables that represent volume of bicyclists and (2) to identify general relationship between the CMFs and these characteristics. In order to achieve better performance of CMFunctions, data mining techniques were used. The results of both before–after and cross-sectional methods show that adding a bike lane on urban arterials has positive safety effects (i.e., CMF < 1) for all crashes and bike crashes. It was found that adding a bike lane is more effective in reducing bike crashes than all crashes. It was also found that the CMFs vary across the sites with different roadway characteristics. In particular, annual average daily traffic (AADT), number of lanes, AADT per lane, median width, bike lane width, and lane width are significant characteristics that affect the variation in safety effects of adding a bike lane. Some socio-economic characteristics such as bike commuter rate and population density also have significant effect on the variation in CMFs. The findings suggest that full CMFunctions showed better model fit than simple CMFuncttions since they account for the heterogeneous effects of multiple roadway and socio-economic characteristics. The proposed CMFunctions provide insights into bike lane design and selection of sites for bike lane installation for reducing crashes. [ABSTRACT FROM AUTHOR]