LSI-R: Level of Service Inventory-Revised
Level of Service Inventory-Revised · Also known as: LSI-R, LSI-R-SV, Andrews-Bonta Risk Assessment
The Level of Service Inventory-Revised (LSI-R) is a 54-item assessment instrument developed by Andrews and Bonta (1995) to measure offender risk level and criminogenic needs (dynamic risk factors related to criminal behavior) in criminal justice populations. It is grounded in the Risk-Need-Responsivity (RNR) model of offender rehabilitation and is widely used in correctional facilities, probation/parole services, and forensic settings to inform release decisions, supervision intensity, treatment prioritization, and rehabilitation planning.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
LSI-R is used across criminal justice system stages: risk assessment at sentencing, supervision classification (probation/parole intensity), institutional management classification in prisons, program placement, and release planning. It is applicable to adult offenders; validity for adolescents is limited (separate LSI-Youth tool exists). LSI-R is particularly valuable for its dynamic need dimensions, which differentiate it from static-only tools and facilitate treatment planning. It is used by probation, parole, institutional correction, and community supervision agencies.
Strengths & limitations
- Integrates risk and need assessment; measures both prediction risk and changeable need factors guiding treatment targeting.
- Grounded in empirically validated RNR model; predictive validity demonstrated across multiple criminal justice samples and populations.
- Comprehensive domain coverage (10 domains) captures multiple criminogenic factors (employment, substance abuse, family, attitudes, peers, housing) relevant to offending.
- Practical utility for case management; domain scores directly guide supervision intensity, program placement, and intervention prioritization.
- Moderately brief (54 items, 45–90 minutes) compared to comprehensive forensic assessments; suitable for high-volume probation/parole settings.
- Extensive research validation in criminal justice; LSI-R is among the most studied risk assessment tools and demonstrates solid predictive validity for recidivism.
- Supported by criminal justice policy and practice standards; widely recommended in correctional best-practice guidelines and evidence-based sentencing.
- Primarily focuses on criminal/general recidivism; less specifically validated for violent recidivism prediction (specialized instruments like HCR-20 or VRAG are stronger for violence risk).
- Dependent on interview quality and honesty; offenders may minimize or deny criminal attitudes, substance abuse, or relationship problems. File corroboration is essential but may be incomplete.
- Domains vary in item count and scoring; some domains (e.g., Financial with 2 items) are less robust than others (e.g., Criminal History with 4 items, or Attitude/Orientation with 7). Domain score reliability differs.
- Scoring partly subjective; interpretations of 'inadequate' housing, 'poor' money management, or 'unstable' relationships may vary across assessors, reducing inter-rater reliability.
- Limited incorporation of protective factors and strengths; LSI-R focuses on risk/need deficits, not assets. Instruments like SAPROF complement LSI-R by measuring protective factors.
- Static items (Criminal History, age at first offense) cannot change; LSI-R total score includes fixed historical factors, limiting its utility for tracking intervention response over time.
- Does not provide probability estimates; yields categorical risk levels (Low/Medium/High) and domain need profiles, not quantified recidivism probability.
Frequently asked
Is LSI-R used to assess violent or sexual recidivism?
LSI-R is primarily validated for general criminal recidivism. While it includes some violent offense history items, specialized instruments (HCR-20 for violence risk, Static-99R for sexual recidivism) are stronger predictors of specific offense types. LSI-R may be used as part of comprehensive violence risk assessment but should not be sole instrument for violent recidivism prediction.
Can LSI-R scores decrease if an offender makes positive changes?
Partially. Static domain items (Criminal History, Age First Offense) do not change. However, dynamic domain scores (Employment, Substance Abuse, Family, Attitudes, Companions, Accommodation, Leisure) can improve with intervention. LSI-R reassessment after treatment or life changes documents improvement in modifiable factors, though total LSI-R score will remain elevated due to fixed criminal history.
How often should LSI-R be readministered?
LSI-R is typically administered at intake (beginning of probation/parole or program entry) and reassessed periodically (annually or at major life events: job change, relationship change, new offense, program completion). More frequent reassessment is unnecessary because dynamic change occurs gradually. Reassessment frequency varies by jurisdiction and case management protocol.
Can offenders refuse to complete the LSI-R?
In most criminal justice settings, LSI-R is mandatory assessment. Refusal or evasiveness is noted in the assessment report and may affect scores (e.g., missing information on attitudes or substance use). LSI-R can be scored based on file review alone if interview is refused, though interview data are preferred for domain accuracy.
What's the difference between LSI-R and VRAG?
LSI-R is a general criminal risk and need assessment tool applicable to all offender types; VRAG specifically predicts violent recidivism in mentally disordered offenders. LSI-R includes modifiable need dimensions; VRAG is static-only. LSI-R is used for supervision planning; VRAG is specialized violence prediction. Instruments serve complementary roles.
Sources
- Andrews, D. A., & Bonta, J. (1995). The Level of Service Inventory-Revised. Department of Psychology, Carleton University. link ↗
- Bonta, J., Law, M., & Hanson, K. (2007). The prediction of criminal and violent recidivism among mentally disordered offenders: A meta-analysis. Psychological Bulletin, 123(2), 123–142. DOI: 10.1037/0033-2909.123.2.123 ↗
How to cite this page
ScholarGate. (2026, June 3). Level of Service Inventory-Revised. ScholarGate. https://scholargate.app/en/forensic-psychology/level-of-service-inventory
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- HCR-20v3Forensic Psychology↔ compare
- PCL-SVForensic Psychology↔ compare
- SAPROFForensic Psychology↔ compare
- VRAGForensic Psychology↔ compare