Suggestive or misleading interrogation techniques may have the effect that innocent people start to remember having committed a serious crime. Confessions are therefore not the best possible evidence, especially not when it is obvious that the interrogation contained elements of suggestion and deception. The problem is illustrated by a case that has become famous in The Netherlands, because two innocent men were imprisoned for about eight years, after obviously false confessions. The confessions were obtained during long and repeated interrogations in which various types of psychological deception were used. In the end, the amount of contradiction, and even of sheer impossibilities, made it clear that the confessions were false and the men innocent. Some of the literature on the creation of false memories is reviewed. It is argued that the practice of criminal investigation may elicit even stronger effects, because empirical research is constrained by ethical limits. The objective of criminal investigation seems to put no limit on what is deemed acceptable, even though we know quite well that the elicitation of false confessions is a serious risk. European agreements about criminal interrogation techniques may provide an effective protection against undesirable practices; but it will not be easy to convince the European legislators of this.