Newcastle upon Tyne, Hunters Moor Regional Rehabilitation Centre,Newcastle upon Tyne NE2 4NR, U.K., telephone 44 (0)191 219 5661, fax 44 (0) 191 219 5665, e.mail: m.p.barnes@ncl.ac.uk Earlier papers have detailed the conceptualization, development, and testing of the Newcastle Independence Assessment Form (NIAF-R), an alternative global outcome measure. This article establishes the predictive validity of the measure. Predictive validity was examined by applying two different prediction models, using functional status and return to work (RTW) as the two outcome variables. Multiple linear and logistic regression analyses indicated that the NIAF-R, when incorporated in equations with age and injury severity, the latter determined by length of coma, was consistently predicting functional status from as early as 8 weeks and return to work from as early as 12 weeks post-injury. Prediction of functional status was especially significant, with the NIAF-R accounting for up to 90 percent of the variance. Prediction of return to work, although consistent, was less powerful. Implications for the validity of the NIAF-R, with particular reference to prognosis and rehabilitation planning, are presented.In the current climate cO head injury rehabilitation, with pressure coming tu hcar to establish cost-effectiveness of rchahitirarion CrWICIIWI1C, therc is a continued driB'c within rchahilitation research to demonstrated the cfficacy c~f rehahititation for this group. If we are to evaluate intervention or to continue to plan appropriately for an imliviclUal's rehabilitation program or post-discharge care, there will he an 1111I11lItMIUe interest in outcome IllCasurement. The usefulness l7t an outcome measure rests entirely on its clHKeptu,lli:,Hil1n and the strength