Correctional assessment strategies and standards of practice should help to
identify those offenders at time of admission who are "at risk" for admini
strative segregation while In prison. An earlier 1997-study began with a li
sting of federally sentenced offenders in administrative segregation. Corre
ctional Service of Canada's Offender Management System was used to identify
a group of segregated and randomly selected non-segregated offenders for c
omparative purposes. These groups were used to make comparisons on a wide v
ariety of case-specific variables. An abundance of case-specific factors as
sessed at prison Intake were found to set Segregated offenders apart from t
heir non-segregated counterparts. The present study extends earlier work in
comparing segregated and non-segregated offenders in relation to three sel
ected outcome measures, discretionary release (parole), return to federal c
ustody for any reason and return with a new offence while on conditional re
lease. This research also explores the predictive value of three objective
classification instruments used to gather case-specific information on crim
inal history background, release risk and security level designation in rel
ation to the various outcome measures for segregated and nonsegregated offe
nders.