There are few long-term studies of the convictions of persistent offenders and the extent to which they may desist from offending. The Sheffield Desistance Study interviewed 113 men aged 19–22 years over four or so years. Their subsequent convictions over the next ten years generally show a continuing pattern of convictions, but with major crime-free gaps. Initial hopes for desistance, if seen as continuing cessation from crime, have not been fulfilled, linked to both substance abuse and, it is argued, recent penal policy in England and Wales. The question is then how we should see recidivism and desistance.