Purpose This paper aims to examine the relationship between neuropsychological functioning as assessed following admission to a high secure psychiatric care (HSPC) hospital and subsequent time spent in long-term segregation (LTS). Evidence from forensic populations suggests neuropsychological functioning difficulties can increase vulnerability to interpersonal violence. However, the impact of this relationship on restrictive interventions used in these settings is poorly understood. Design/methodology/approach This study quantitatively examined the neuropsychological profiles of 80 male HSPC patients as assessed during routine admission assessments, comparing data against any subsequent LTS duration during a one-year period, and a non-LTS control (n = 27). Findings Analysis found individuals who were willing and able to complete a routine neuropsychological admissions assessment spent significantly less time in LTS than those unable to complete the assessment. Performance within a test of novel problem solving (Key Search Test) was significantly worse in the LTS group than controls. Performance within a visual memory task (Immediate Recall section of the Rey Complex Figure Test) significantly correlated with LTS duration. Additional findings suggest the absence of self-reported planning difficulties as measured by a Dysexecutive Questionnaire (DEX) negatively correlated with LTS duration, while self-reported restlessness was positively correlated with LTS. Practical implications This has implications for early assessment of LTS risk and potential use of cognitive interventions to reduce the use of restrictive practices. Originality/value The results suggest some aspects of neuropsychological performance as assessed during admission to a HSPC hospital appear to be related to subsequent time spent in LTS.