jats:titleAbstract</jats:title> jats:sec jats:titleResearch Question</jats:title> jats:pHow concentrated is the total harm of offences with detected offenders (identified suspects) among the complete list of all detected offenders in a given year in an English police agency, and how consistent is the list of highest-harm “felonious few” offenders from one year to the next?</jats:p> </jats:sec> jats:sec jats:titleData</jats:title> jats:pCharacteristics of 327,566 crimes and 39,545 unique offenders as recorded by Northamptonshire Police in 7 years from 2010 to 2016 provide the basis for this analysis.</jats:p> </jats:sec> jats:sec jats:titleMethods</jats:title> jats:pCrime and offender records were matched to harm weightings derived from the Cambridge Crime Harm Index (Sherman et al. 2016a; Sherman et al., jats:italicPolicing</jats:italic>, jats:italic10</jats:italic>(3), 171–183, 2016b). Descriptive statistics summarize a concentration of harm identifying the felonious few, changes over time in membership of the “few”, offender typologies and tests for escalation of severity, frequency and intermittency across repeated offences.</jats:p> </jats:sec> jats:sec jats:titleFindings</jats:title> jats:pCrime harm is much more concentrated among offenders than crime volume: 80% of crime harm that is identified to an offender is linked to a felonious few of just 7% of all detected offenders. While chronic repeat offenders are the majority contributors to harm totals of this group, those with the most general range of offence types contribute the most harm. Individual members of the felonious few rarely maintain that position year on year; over 95% of each year’s list is composed of individuals not present in previous years. Within individual crime histories, we observe a pattern of de-escalation in crime harm per offence over time. “One-time” offenders, those with just one crime record, typically made up a third of the felonious few in both number and harm contribution.</jats:p> </jats:sec> jats:sec jats:titleConclusions</jats:title> jats:pThese findings demonstrate the potential to target a small number of repeat offenders for harm reduction strategies using a metric of total crime severity, not just volume, despite a substantial portion of crime harm caused by one-time offenders that may be largely unpredictable.</jats:p> </jats:sec>