Many studies have examined the relationships between students' perform
ances on the National Board of Medical Examiners (NBME) Part I and Par
t II examinations and their postgraduate clinical performances. Most s
tudies have found a positive relationship between students' NBME exami
nation scores and the ratings of residents' clinical performances and/
or scores on specialty board certification examinations. Surveys of re
sidency directors have also shown that NBME scores are used as part of
the process for selecting residents, although other data and other se
lection criteria are considered more useful than the NBME scores. One
area of continued concern is that a large body of data on the United S
tates Medical Licensing Examination (USMLE) is not yet available. Howe
ver, the predictive validity of the NBME scores supports the use of US
MLE scores as part of the review process to screen potential residents
. Residency directors should continue to use a variety of criteria in
their final selection decisions and are encouraged to use their own pr
ogram-specific data to establish and monitor particular cutoff scores
for screening applicants.