Crime statistics attempt to provide a statistical measure of the level, or amount, of- crime that is prevalent in societies. Given that crime, by definition, is an illegal activity, every way of measuring it is likely to be inaccurate.

There are many forms of measuring crime, including household surveys or checking hospital or insurance records, but the term "crime statistic" usually refers to figures compiled by the Police and similar law enforcement agencies. However, it is well known that many if not most offences are not reported to the police, and changes in police procedures can have a big impact on how such reported crimes are categorised. This is why public surveys are sometimes conducted to estimate the amount of crime not reported to police and to ascertain levels of victimization. Such surveys are usually more reliable in providing reliable trends over time, but they rarely encompass all crime (eg separate surveys are required to measure retail crime such as shoplifting, as distinct from crime against the public), rarely give local statistics useful for local crime prevention or enforcement, often ignore offences against children, and do not count offenders brought before the criminal justice system.

Crime statistics are gathered and reported by many countries and are of interest to several international organizations, including Interpol and the United Nations. Law enforcement agencies in some countries, such as the FBI in the United States and the Home Office in England & Wales, publish crime indices, which are compilations of statistics for various types of crime.

Statistics are usually collected on

• Offences - Breaches of the law
• Offenders - Those who commit offences
• Victims - Those who are offended against

Because laws vary between jurisdictions, comparing crime statistics between, and even within, countries can be difficult, sometimes even problematic. Even overall international trends can be difficult to interpret authoritatively. Research using a series of victim surveys in 18 countries of the European Union funded by the European Commission has reported (2005) that the level of crime in Europe has fallen back to the levels of 1990, and notes that levels of common crime have shown declining trends in the U.S.A., Canada, Australia and other industrialised countries as well. The European researchers say a general consensus identifies demographic change as the leading cause for this international trend. However they suggest that "increased use of crime prevention measures may indeed be the common factor behind the near universal decrease in overall levels of crime in the Western world", since decreases have been most pronounced in property crime and less so, if at all, in contact crimes.

Crime indexes are generated to analyze crime statistics. The most common source of such statistics are records reported to the police. Critics do not feel that crime indexes are particularly useful as they only list reported crimes. This makes it difficult to know whether a "10% increase" has any real meaning. An alternative source of data is the victimization survey, usually conducted via annual interviews with households, whose occupants are asked whether they have been the victim of crime. Critics note that surveys are generally unreliable, especially when they involve discussions of criminal activity. Victims surveys however provide more reliable data than police recorded crime for the volume (or common) crimes. Rare but serious crimes such as intentional homicide and armed robberies have frequencies that are too low to be measured accurately by a survey. Considering the seriousness of these crimes and the amount work done on them, police statistics are more accurate here, at least in the developed countries.