Kempster, R v

Expert ear-print comparison evidence from an experienced examiner may be relevant and admissible. Its probative value depends on the quality and precision of the particular comparison.

Where only gross ear features are available, a positive identification can safely be drawn only if those features provide a precise match. A comparison which establishes only that a mark is consistent with the defendant's ear, while showing material mismatch, cannot by itself support a conviction where the jury were required to be sure of identity.

The quashing of one count does not undermine convictions on other counts where the jury were clearly directed to consider them separately and the remaining evidence independently supports them.

The appellant had been convicted of three burglaries and attempted burglary. Following an earlier unsuccessful appeal, the Criminal Cases Review Commission referred the conviction on the first burglary count after new expert evidence challenged the reliability of the ear-print comparison used at trial.

The court had already rejected a separate argument concerning an alternative verdict on the fourth count. It heard fresh evidence from a mathematician involved in research into ear-print comparison and from the prosecution ear-print examiner. The appellant also sought leave to challenge the remaining convictions on the basis that an unsafe verdict on the first count may have affected the jury's assessment of the other counts.

The central issues were whether the ear-print evidence could safely establish identity on the first count, and whether any resulting defect affected the other convictions or the concurrent sentence.

1. Appeal allowed in part. The court quashed the conviction on count 1. It refused leave to pursue the further challenge to counts 2, 3 and 4, and left both those convictions and the concurrent sentence of 10 years' imprisonment undisturbed.

2. Ear-print evidence from an experienced examiner was capable of being relevant and admissible. The decisive question was its probative value in the individual case. Ear comparison could potentially permit positive identification, particularly where matching minutiae were available. Where the comparison depended on gross features alone, however, a positive match could safely be found only if those features provided a truly precise match.

3. The fresh evidence showed a close general similarity between the scene mark and the appellant's ear, including the notch and nodule. It established consistency with the appellant having left the mark. It did not establish a precise match. The mismatches might be explicable by pressure or movement, but were too extensive for the comparison alone to justify a guilty verdict. Since the jury had been directed to acquit unless sure that the ear-print was the appellant's, the conviction on count 1 was unsafe.

4. The verdict on count 1 did not supply material for convictions on counts 2, 3 and 4. The summing-up had clearly required separate consideration of those counts and made counts 2 and 3 depend upon proof of count 4, rather than upon count 1. The evidence concerning the stolen credit card and the account given by Robert Smith independently supported the remaining verdicts.

5. The remaining offences, committed against vulnerable elderly victims and viewed with the appellant's relevant antecedents, independently justified the severe concurrent sentence. No reduction was warranted after count 1 was quashed.

• Court of Appeal (Criminal Division): On a reference by the Criminal Cases Review Commission, the court quashed the conviction on count 1 but refused leave to challenge counts 2, 3 and 4 on the fresh-evidence ground. It did not alter the sentence.

• Court of Appeal (Criminal Division): On 11 December 2003, the court dismissed the earlier appeals against conviction and sentence.

• Single judge: Leave to appeal against sentence was granted, but leave to appeal against conviction was initially refused. The conviction application was subsequently renewed and leave was granted by the full court.