Results found that most distances examined were short. In homicides, distances were below 1 kilometer, in rapes below 2 kilometers, whereas in robberies against businesses almost 6 kilometers. Some crime features were correlated with the distances in rape and homicide cases. Combining the behavioral information to a spatial behavior measure allowed for better prediction of travelled distances compared to using single variables. This work sought to describe distances from home to offense locations (journey-to-crime) of offenders in difficult-to-solve homicides and rapes as well as robberies against businesses; determine whether the distances in these offenses differed from each other; and test whether selected features related to the offense would be associated with the distances. The results were said to have practical implications for crime investigations as the crime features explored were, as a rule, such that they would be known by the police prior to the offender being identified. It is noted that a general theoretical framework for binding together journey-to-crime distances and offender crime scene behavior and other important crime features is needed. Lists of 99 difficult-to-solve homicides, 56 rapes, and 275 robberies against businesses from the city of Milan, Italy, were acquired. The collected data consisted of home and offense location coordinates with information on the behavior of the offenders (for rapes and homicides). The journey-to-crime functions were calibrated using the journey-to-crime module of CrimeStatIII©. Tables, figure, and references