There are increasing concerns about the use of neuropletics in the treatment of challenging behaviours. Their use has been associated with cerebrovascular events, falls, weight gain, accelerated cognitive decline, while showing limited evidence of efficacy (Sink et al., 2005). In the light of such concerns, Howard and colleagues ' (2001) guidelines suggest that non-pharmacological interventions should be used prior to the use of pharmacological treatments. Turner (2005) argues that effective non-pharmacological interventions, for individuals displaying challenging behaviour, should be based on a multifaceted approach to assessment. She suggests that one employs applied behaviour analytic techniques within clinical formulations, such as conducting a functional analysis of people's challenging behaviour.A clinical objective when investigating the nature of challenging behaviour via a functional analysis is to obtain useful information from staff regarding the triggers and setting conditions associated with the behaviour. The present work evaluates the level of agreement of staff regarding triggers for residents' behaviour.This study is based on clinical work undertaken by the Newcastle Challenging Behaviour Team (NCBT). The current assessment procedures for the NCBT are based on a formulation-led model, incorporating features of cognitive behaviour therapy and the unmet needs framework. Staff are seen as key stakeholders in the treatment strategy of the NCBT (James et al., 2001), and they are consulted frequently providing relevant information about the nature of the referred patient's challenging behaviour.In the present study, 76 care staff acted as informants. Twelve (15.8%) staff were qualified, 50 (65.8%) were unqualified and the remainder did not provide their status. The patients were a convenience clinical sample of older adult service users (n ¼ 38, mean age ¼ 81.4 years, SD ¼ 7.40, referred to NCBT for treatment of their challenging behaviours; 76.3% of the sample came from care homes and 23.7% were inpatients at the local hospital; 73.5% of patients were diagnosed with a specific dementia.For each patient, two staff were asked to indicate the situations and settings that triggered the challenging behaviours, using a Functional Analysis Checklist (FAC). The FAC is an unpublished measure based on previous informant-based methods of functional analysis (e.g. Zarkowska and Clements, 1994).