Six police forces participated in the study; data were collected from a total of 11 stations. Documentary data about incidents leading to section 5 arrests during 1990 were extracted from case files. Interviews were conducted with a range of officers at each station, and observation of policing in each area was conducted during peak times at which disorder could be expected to occur. The report raises questions about the extent of police intervention in incidents of low-level disorder and whether it is always appropriate to use section 5 to make arrests. First, although the behavior leading to section 5 arrests is often offensive, it may not be sufficiently serious to offend the members of the public present or the police themselves. Secondly, although there are obvious difficulties in reviewing police decisions after the event, there are grounds for questioning whether arrest or use of section 5 was always the right course of action. If, on the one hand, the objective of arrest was to instill respect for the police, this probably was not achieved. If, on the other hand, the aim was to remove disorderly persons from the streets, the report questions whether it was necessary to use section 5 as the mechanism. Section 5 will be more useful if it is reserved for those cases for which it is intended: those in which vulnerable persons are genuinely likely to, or do, suffer from offensive behavior. 8 tables and 20 references