PURPOSE: To examine the validity and reliability of a modified Reintegration to Normal Living Index (mRNL Index) with a sample of community-dwelling adults with mixed diagnoses. METHOD: Forty-six adults (mean 55.2 ± 20.3 years) were recruited through convenience sampling from outpatient rehabilitation services. They completed the mRNL Index, Community Integration Measure (CIM) and Life Space Assessment (LSA) and were invited to complete them again 2 weeks later. RESULTS: Construct validity of the mRNL Index was confirmed by good fit to the Rasch measurement model. The mRNL Index demonstrated acceptable internal consistency (Cronbach's α = 0.80), as did the Daily Functioning subscale (Cronbach's α = 0.80) and Personal Integration subscale (Cronbach's α = 0.82). Test-retest reliability was also acceptable (intraclass correlation coefficient (3,1) = 0.83, p = .0001). As hypothesised, the LSA did not correlate with the Personal Integration subscale (Spearman rho = 0.08) and moderately correlated to the Daily Functioning subscale (Spearman rho = 0.59). The CIM was moderately correlated with the Personal Integration subscales (Spearman rho = 0.54) and the Daily Functioning subscale (Spearman rho = 0.53), though higher correlation was expected with the latter. CONCLUSION: Modifications to the phrasing, rating scale and subscales improved the validity of the original RNL Index for a mixed rehabilitation, community-dwelling population.