JUDGES WHO PRIDE THEMSELVES ON THEIR LACK OF TECHNICAL KNOWLEDGE ARE RAPIDLY BECOMING EXTINCT. CASES IN WHICH MUCH SCIENTIFIC EVIDENCE IS EXPECTED SHOULD BE TRIED BY A PANEL OF THREE TO FIVE JUDGES WHO CAN POOL THEIR SCIENTIFIC KNOWLEDGE. THE JUDGES' FREEDOM TO EVALUATE EVIDENCE FINDS ITS LIMITS WHERE SCIENTIFIC LAWS ESTABLISH RELIABLE GUIDELINES (E.G., WITH BLOOD TESTS IN PATERNITY SUITS). IN IDEAL CASES, DIRECT WITNESS TESTIMONY AND DEMONSTRATIVE PROOF COMPLEMENT AND SUPPORT EACH OTHER. DESPITE ITS OBVIOUS SHORTCOMINGS AND UNRELIABILITY, DIRECT WITNESS TESTIMONY STILL RECEIVES A MAXIMUM OF PUBLIC ATTENTION AND MEDIA COVERAGE, WHILE WEIGHTY DEMONSTRATIVE EVIDENCE FREQUENTLY GOES UNREPORTED. EVEN WITH DEMONSTRATIVE EVIDENCE, THE HUMAN FACTOR IS OF THE UTMOST IMPORTANCE: TRAINED PERSONNEL MUST EXERCISE THE UTMOST CAUTION AND ACCURACY IN DISCOVERING, LABELING, AND PRESERVING THE TRACES OF A CRIME. THE JUDGE HIMSELF MUST ALWAYS ALLOW FOR HUMAN ERROR; ALL EXPERT WITNESSES ARE ULTIMATELY WITNESSES, ALTHOUGH HIGHLY QUALIFIED ONES. IN MANY CASES, A PIECE OF EVIDENCE--THOUGH RELAIBLE--DOES NOT POINT TO A PARTICULAR OFFENDER (OR TO SEVERAL OFFENDERS), AND A CAREFUL SYNTHESIS OF THE ENTIRE INVESTIGATIVE PROCESS IS REQUIRED TO DETERMINE THE TRUE OFFENDER. IN SUMMARY, SCIENTIFIC PROOF OFFERS GREAT OPPORTUNITIES TO CRIMINAL JUSTICE IF JUDGES ACCEPT ITS CHALLENGES AND RECOGNIZE ITS LIMITATIONS. THE ARTICLE CONTAINS NO REFERENCES. --IN GERMAN. (SAJ)