Serial offending and co-offending are two of the most prevalent offending behaviors in society. Evidence increasingly suggests that serial offending and co-offending are interdependent. However, extant research has generally studied both offending behaviors separately, misrepresenting the nature and extent of serial and co-offending and their potential interdependencies. To resolve this, we leverage the recent availability of forensic biometric data to crime researchers. This allows us to uniquely distinguish offenders and link offences and co-offenders across space and time, without the need for offenders having been identified by the police. By integrating police data and forensic biometric data into a single robust crime dataset we are able to study the serial and co-offending behaviors of identified and unidentified offenders—which is not possible when only using police data. Network theory offers a holistic perspective of offending behavior by representing serial offending and co-offending in a single dynamic network that evolves in space and time. ABM is a computational method that allows to simulate interactions between offenders and within offender groups via simple behavioral rules that are rooted in real-world observations obtained from our integrated dataset. Combining the outcomes from the network analysis and ABM, we generate a quantitative behavioral framework on serial and co-offending behaviors.