Researchers from the University of Florida (UF) and Florida International University (FIU) are developing predictive methods to determine crash frequency on freeways with HOV and HOT lanes. Consistent with the procedures outlined in the Highway Safety Manual (HSM), Dr. Siva Srinivasan of UF and Dr. Albert Gan of FIU are working on models that will include safety performance functions (SPFs) that relate the crash frequency to traffic volumes and segment lengths and crash modification factors (CMFs) to control for roadway geometry features and the types of separation between the general purpose and the HOV/HOT lanes. Separate equations are estimated by crash severity (one equation for total crashes and one for fatal and injury crashes). Data from the states of California, Washington, and Florida have been assembled to develop models for facilities with HOV lanes. Collectively, these represent 5-years of data for over 300 miles of facilities with HOV lanes. Data from California, Minnesota, Texas, and Florida are being assembled to develop the models for facilities with HOT lanes. The initial data represents about 25 miles of facility and approximately 5 years’ worth of data for each facility. The models will be implemented in a simple spreadsheet tool to allow analysts to use the equations for predictive assessments. Overall, the research will provide procedures that will help FDOT consider safety in decisions about planning and designing freeways with HOV or HOT lanes.