The sample drawn for police stations in West Yorkshire, Nottinghamshire, Avon and Somerset, and London consisted of 218 suspects and covered a wide range of criminal offenses, albeit no exceptionally serious ones. Nearly half the suspects interviewed at the police stations made a confession, and over two-thirds gave information that would help secure a conviction. Almost one-fifth volunteered information about offenses other than those for which they had been detained. However, in only 8 percent of the cases did the officers interviewed say they would have dropped the case if the suspect had refused to answer questions, while in 56 percent they would have relied on the evidence already available. A few apparent breaches of the Rules occurred, such as interviewing a juvenile without the presence of a suitable adult and not cautioning a suspect before questioning. The overall fairness of police practice and the absence of any incident likely to impair the reliability of prosecution evidence suggest that the police could be expected to deal with suspects in accordance with realistic rules. Observers felt that confessions were often made because they were an easy and logical way out of a tense and uncomfortable situation created by persistent and determined officers. Thus, if more weight were to be given to the principles of voluntariness and right to silence in the Rules, it probably would be necessary to reduce police questioning powers. If police interrogation is sanctioned, the present Rules may be unnecessarily strict and would be followed better if they were relaxed. Tables, footnotes, and five references are included.