Approved For Release 2004/SUC     RDP96-00788R001000340002-3
 THE JOINT CHIEFS OF STAFF
 WASHINGTON, D.C. 20301
 11 February 1981
 MEMORANDUM FOR COMMANDER USA INSCOM
 Through:   USA/ACSI
 Subject:   Grill Flame Evaluation (U)
 IRA At
 1.  (SNFD) During the period 23 November 1979 through 13 January
 1981, USAINSCOM/ADCSOPS-HUMINT/Special Action Office provided
 202 Grill Flame Reports to OJCS/J3/SOD in support of the Iranian
 Hostage situation.  Of this total 118 reports were provided prior
 to 24 April 1980 and 84 subsequent to that date.
 2.  (SNFD) Source targeting was divided between personalities and
 physical locations as follows:
 Phase I
 - Personalities
 68
 Locations
 50
 Phase II - Personalities
 52
 Locations
 32
 3.  (SNFD) Comparison of these reports with returnee debriefings
 revealed a very low correlation between actual hostage locations/
 conditions and inferences drawn from Grill Reporting.   Only
 seven reports could be positively correlated with actual location
 or condition.  Approximately 59 reports revealed a possible or
 partial correlation.  However, these same reports often included
 erroneous data.   Sixteen reports contained inconclusive data
 making correlation highly subjective.              Eight reports were noted
 as being poor from an administrative/procedural standpoint and
 therefore being of no value.  One hundred and twelve reports were
 found to be entirely incorrect.
 ARMY and JCS review(s) completed.
 CLASSIFIED BY:  Chief, SOD
 DATE FOR (  ) DECLASSIFICATI
 OR (X) REVIEW
 IS:       11 Feb 2001
 EXTENDED BY:, Chief, SOD
 REASON:  5200.1R, Para 2-30lc5&6
 Approved For Release 2004/OQi617DP96-00788 R001000340002-3
 Approved For Release 2004/0DP96-00788 R001000340002-3
 4.  (SNFD) Although the above comparison would appear to degrade
 the value of the Grill Flame program, several factors must be
 considered before drawing any final judgements.   Some of these
 factors are:
 a. Contamination of Grill Flame reporters could not be
 avoided because of the extensive media and intelligence
 treatment accorded the incident.
 b. Effective employment of Grill Flame concepts may be, at
 least partially, dependent upon the cooperation of the HUMAN
 subject and reliable recognition features of the physical/
 structural targets.
 c. The lack of reliable and objective analysis techniques is
 also a major limiting factor in evaluating Grill Flame Reporting.
 As an example, midway through the reporting period this office
 requested the Project Manager to conduct several sessions on a
 special subject.   Two sessions were run on a controlled basis,
 i.e., the actual location and activities of subject were known
 to this office but not to the Project Manager.  Evaluation of
 the reports and follow-on questioning of the subject revealed
 that reports could not be accurately interpreted until the
 actual physical and mental activities of the subject were known.
 Once these became known, a greater degree of correlation occurred.
 5.  (SNFD) In summary Grill Flame Reporting, as evidenced by the
 above results, is not yet proven to be reliable enough to use in
 a tactical/operational environment.  The Grill Flame concept may
 hold future potential if positive controls and effective analytical
 techniques are developed.  Recommend any future Grill Flame efforts
 be conducted under extremely controlled conditions and care be taken
 by evaluators and customers to avoid becoming enamoured of the source.
 R. J .ENAHAN
 LTCOL, USAF
 Approved For Release 2004/0
 1DP96-00788R001 000340002-3