Abstract Little is known about how the Sentencing Council’s guidance to treat intoxication as aggravation is applied in practice. With reference to assault offences, this study examines: whether intoxication has an aggravating effect; whether this is moderated through other characteristics of the case; and whether any effect is consistent across Crown Court locations. The probability of custody and sentence severity are modelled using (ordered) logit multilevel models and data from the Crown Court Sentencing Survey. The probability of receiving a custodial or severe sentence when intoxication features is increased, however, is moderated if the offence is deemed an isolated incident. Effects are relatively consistent across Crown Court locations, however ongoing monitoring of how intoxication shapes sentencing practice it is encouraged.