Material for the analysis was based on government pronouncements, policy statements, consultative documents, and parliamentary debates; semistructured interviews were conducted with 28 families and with 35 professional personnel, including police officers from five local stations. The findings show that in consultations with police about their children's glue- sniffing, mothers often experienced difficulties. Police often were indifferent to the information that mothers provided about their own problems and conditions, contributing to mothers' frustration and anger. The author argues that what happened to mothers and children can be understood as the outcome of the decisionmaking processes of the police, of institutional constraints and organizational priorities, and of the administrative need to fit children and their families into a restricted and rigid number of categories provided by professional discourse. Since glue-sniffing can be viewed as both a manifestation of normality and a symptom of pathology, this creates the possibility of some children being defined as nonproblematic by the police. These discourses can be drawn on to legitimate their inaction in relation to children with no other "offenses" or "problematic" personal circumstances apart from glue-sniffing; however, the same discursive processes can be drawn on to construct other children as "too problematic," and the notion of "treatability" may be invoked to legitimate nonintervention in relation to children who do manifest these problems. Those children considered "manageable" are included for intervention and referred to other professionals. These, however, are not necessarily the chronic cases envisioned by the government. 36 references