Crash prediction models, also known as safety performance functions (SPFs), were developed in this study for different functional classes of roadways. SPFs were developed for total crashes and equivalent property-damage-only (EPDO) crashes to investigate the effects of pavement friction, roadway geometry, road surface conditions, and vehicle body types on traffic crash frequencies. Crash data, roadway geometric design data, and friction data obtained from the Wyoming Department of Transportation (WYDOT) were processed to develop the SPFs. The results of the SPFs revealed that lower friction numbers not only increased the occurrences of total crashes but also increased the occurrences of EPDO crashes. This suggests that maintaining adequate friction levels reduces both total and EPDO crashes. Several geometric characteristics of roadways were found to be significant from the results of the SPFs. Upgrades and downgrades were found to increase EPDO crashes in most facilities. Tangent segments were found to reduce total crashes but increase EPDO crashes. When road surface condition was considered, it was found that dry and wet road surfaces were positively associated with both total and EPDO crashes. The average annual daily traffic (AADT) volume, an exposure measure, was positively associated with the frequencies of total crashes and EPDO crashes, assuming that all else was unchanged. The results from the SPFs are of great value and will be beneficial in identify locations that are at risk of experiencing crashes because of low pavement friction numbers.