A U S T R A L I A N I N S T I T U T E
O F C R I M I N O L O G Y
t r e n d s
&
i s s u e s
No. 226
DNA Identification in the
Criminal Justice System
Jeremy Gans and Gregor Urbas
in crime and criminal justice
May 2002
ISSN 0817-8542
ISBN 0 642 24262 3
Australian Institute
of Criminology
GPO Box 2944
Canberra ACT 2601
Australia
Tel: 02 6260 9221
Fax: 02 6260 9201
For a complete list and the full text of the
papers in the Trends and Issues in
Crime and Criminal Justice series, visit
the AIC web site at:
http://www.aic.gov.au
Disclaimer: This research paper does not
necessarily reflect the policy position of the
Commonwealth Government.
Adam Graycar
Director
DNA profiling and the forensic use of DNA evidence have undergone
considerable development since the Australian Institute of Criminology first
examined this topic in 1990 in Trends and Issues no. 26. Some of the
laboratory techniques described in that report have since been refined so that
more precise DNA profiling is now possible, and a greater range of criminal
investigations can benefit from the use of such forensic techniques. Moreover,
the proposal in that report for a national DNA database has now been
advanced, with the establishment on 1 July 2000 of the CrimTrac agency.
However, many of the issues raised in relation to scientific reliability,
standardisation of profiling techniques, laboratory accreditation and quality
control, improved population and data analysis, and privacy are still the
subject of disputation in legal proceedings.
This paper examines the science of DNA identification and its use during
criminal investigations and in criminal proceedings, including criminal trials,
appeals and post-conviction proceedings. It describes the main benefits and
costs of the increasing role of DNA identification in the criminal justice system.
The Science of DNA Identification
A Natural Identifier
Deoxyribonucleic acid (DNA) is a long molecule, found in the cellular
nuclei of living organisms. Since 1954, scientists have recognised that the
chemical structure of an individual’s DNA encodes information about that
individual’s inherited characteristics. The present limits on genetic science
mean that a direct analysis of a person’s DNA will yield only limited
information about individual characteristics, although some research
suggests that investigators may in the future be able to discern specific
physical traits such as hair, eye and skin colour from forensic samples
(National Institute of Justice 2000, pp. 18–19; van Oorschot et al. 2001).
Rather, the current utility of DNA analysis to the criminal justice system
arises from the comparison of DNA from two sources, such as DNA from
a crime scene and DNA from a suspect, to determine the relationship
between those sources.
Traditionally, the identification of a person has required the observation
of that person’s entire body or of localised special characteristics such as
fingerprints, blood group or hair type. By contrast, DNA analysis allows
identification by reference to the information contained in any human
nucleic cell, irrespective of which part of the body the cell comes from. The
DNA in a human cell is unique, the product of sexual reproduction that
combines half of the mother’s DNA and half of the father’s DNA. Every
cell in an individual’s body is the result of cellular division, which copies
the DNA in the newly fertilised cell into every other nucleic cell. As a result,
DNA in a cellular nucleus is identical throughout a human body but
variable between any two humans, making it a natural alternative to
artificial human identifiers, such as names or tax-file numbers. The notable
exception is identical twins, who develop from a single fertilised cell and
hence have identical nuclear DNA.
The technique of “DNA identification” compares the DNA of two
bodily samples to ascertain whether or not they came from the same human
being. Identity of DNA in the cells across both samples implies that the
samples are derived from the same person (or identical twins); non-identity
implies different human sources. Alternative comparative techniques can

Australian Institute of Criminology
2
be used to determine whether or not
there is a familial relationship between
the two human sources. For example,
a matrilineal relationship can be
inferred from a comparison of DNA
in mitochondria, which pass from
mother to child unchanged by sexual
reproduction.
DNA Profiling
Comparison of human DNA molecules
does not require analysis of the entire
DNA molecule, as about 99.9 per cent
of DNA is common to all people.
Rather, DNA comparison need only
focus on a portion of the remaining
0.1 per cent of human DNA that is
sufficiently variable to be unique to
individuals. Such variable DNA—
termed “non-coding” (or “junk”)
DNA—plays no direct role in the
development of human
characteristics (Trent 2000, p. 52).
Modern comparative techniques
compare only a small set of features
of non-coding DNA. Such sets of
features are known as DNA profiles
and can be represented as an ordered
series of numbers. That DNA profiles
are easily quantified represents a
further advantage over other unique
human features, such as appearance
and fingerprints, as it allows for
automated analysis. The features
comprised in a DNA profile must be
sufficiently variable throughout the
population to have an acceptable
statistical likelihood that the profile is
unique in that population, but also
sufficiently regular to be amenable to
cheap and efficient mass analysis.
While several varieties of DNA
profiling have been used in the past
(National Institute of Justice 2000;
Butler & Becker 2001; Freckelton &
Selby 2002, ch. 80), the future of DNA
identification in Australia is likely to be
dominated by the type of profiling in
present use (Box 1). Any significant
future changes in profiling would
render contemporary investigative
databases obsolete.
Laboratory technicians do not
“read” a DNA profile from a bodily
sample. Rather, they construct a
profile by inference from the outcomes
of a series of procedures performed on
that sample. Contemporary profiling
techniques (Box 1) are increasingly
automated, but the elimination of
artefacts of the profiling process
requires careful judgments by properly
trained scientists (Roberts 1998, p. 32;
Kaye & Sensabaugh 2000, pp. 516–17.)
Accordingly, a DNA profile generated
from a sample by contemporary
procedures must be understood not
as a fact about a sample but rather as
an interpretation of that sample.
Future developments may allow initial
profiling to be done by non-technicians
outside of the lab (National Institute
of Justice 2000, p. 30).
An attempted comparison (or
“matching”) of two DNA profiles in
order to determine whether they are
related will yield one of three possible
results (Table 1).
DNA Identification in Criminal
Investigations
Linking People and Crimes
Crime investigators utilise DNA
profiles from two sources: human
bodies and small samples of human
bodily material. DNA profiles can be
obtained from human bodies by
analysing samples removed from
those bodies. Forensic procedures
that can be used to obtain such
samples (whether voluntarily or
involuntarily) include blood sampling
by injection, pulling out hair at the
root and taking swabs from inside
the mouth, known as buccal swabs.
In many cases, DNA profiles can
be obtained from bodily samples that
have become separated from a human
body. Contemporary profiling
techniques can generally be used on
such tiny samples as the root of a
pulled hair, saliva on a cigarette butt,
a square-centimetre blood stain, skin
cells from clothing or three micrograms
of semen from a vaginal swab;
standard or alternative techniques
will sometimes succeed on other, less
optimal, samples such as shed hair or
skin cells from a handled object (Kaye
& Sensabaugh 2000). Investigators
will be interested in such samples if
they suspect that they became
separated from a person’s body
(usually either victim or offender) at
the time of the commission of a crime,
thus providing a potential insight
into details of that crime.
The most important use of DNA
identification by crime investigators
is to compare a profile believed to be
from a crime perpetrator (for example,
derived from semen in a rape victim’s
vagina, or blood, hair or skin cells at
a crime scene or on a victim’s body)
with a known person’s profile. Other
uses of DNA identification include:
 comparing a profile from foreign
samples on a suspect’s body or
possessions with a victim’s profile
(to test the suspect’s prior contact
with the victim);
 comparing a profile from an
unidentified person or corpse with
a known person’s profile (to test
identity); or
 comparing profiles in two crime
scene samples (to infer the details
of a crime or the common
involvement of one person in
separate crimes.)
DNA matching can be used at various
stages of an investigation. If a known
person is a suspect at the time of the
matching, then a positive match
between crime scene DNA and that
person will help to confirm the
existing suspicion, while a negative
match will tend to negate that
suspicion. However, DNA matching
can also be used before suspicion has
fallen on a single individual by
comparing the unknown sample
profile to samples taken from a group
of persons, such as all adult males
within a locality. A positive match
Box 1: DNA profiling in contemporary Australian forensic laboratories
Since January 1999, all Australian forensic laboratories regularly involved in criminal
casework have used a commercial profiling kit, Profiler Plus, owned by US-based
Perkin Elmer Corporation. Profiler Plus analyses nine points in human DNA where
short sequences of proteins are repeated a variable number of times. The profile consists
of the number of repetitions at each point. For a sample from a single individual there
will be up to two numbers at each of the nine points, one from each parent. “Mixed”
samples from more than one individual (for example, a post-coital vaginal swab) will
produce a more complex profile, requiring considerable interpretation. This general
approach to profiling based on “short tandem repeats” (STRs), which has largely
replaced “restriction fragment length polymorphism”, or RFLP, is favoured in
contemporary forensic laboratories because it is amenable to the analysis of small,
degraded and mixed samples typically present at crime scenes. However, other
techniques can be used in individual investigations where appropriate. The Profiler Plus
system can theoretically distinguish over 10 billion possible variations of human DNA.
The key steps of the analysis are the introduction of fluorescent primers that attach
to the beginning and end of the nine repeating portions of DNA, the simulation of the
natural process of replication at those portions (using the “polymerase chain reaction”,
or PCR, method), and the passage of an electric current that separates the results by
length. The resulting pattern of fluorescent primers can then be visually examined to
discern the number of repeating portions at each analysed point of DNA.