Practitioners have argued that summoning is a more effective and less costly way of bringing offenders before the court than arrest, particularly in minor violations. This study collected data on all offenses for which the power of arrest exists from four divisions of the Merseyside, North Wales, Lancashire, and South Yorkshire Police for October 1978. Analysis conformed unpublished Home Office statistics by revealing considerable variation among police divisions in the proportions of suspects handled by arrest and summons. The division differed much less in the proportion of suspects brought to court whom the police detained in custody, only about one-quarter of all adult offenders. The use of the charge procedure allows the police to take shortcuts in obvious and straightforward cases where guilty pleas are expected and is consequently cheaper than summons in these circumstances. In forces which use summonses extensively, the police themselves exercise a measure of discretion through the formal caution procedure in dealing with less serious crimes. The use of the formal caution was significant for juveniles in all areas. The suspected offender who proceeds to court by summons can expect an 80-day wait for the first court appearance, three times the period experienced by the counterpart handled by charge and bail. The study suggested that considerable cost increases might be associated with expanded use of summoning, that such costs are difficult to predict, and proposed modifications in the summons procedure probably would not reduce these costs significantly. Tables, organizational charts, and other materials supporting the statistical analyses are provided.