Crime in the Metropolitan Police area

By type of offence

Recorded offences in the 12 months to June 2026, with the change on the 12 months before.

Set against the national mix of offences, theft from the person stands out in the Metropolitan Police area: it accounts for roughly 4 times the share of recorded crime it does nationally, with robbery and bicycle theft also over-represented. In the other direction, other crime, criminal damage and arson, and violence and sexual offences all make up well under their national share.

Home Office recorded crime, data.police.uk. Rates use ONS mid-year population estimates (unknown residents). Categories are the Home Office crime types as published; they are not comparable with court or conviction figures.

What happened to those crimes

Police recorded 895,050 offences other than anti-social behaviour in Metropolitan Police Service in the year to June 2026. By the time this data was published, about 49% had been closed with no suspect identified and 25% closed because a prosecution was not possible or not pursued. Around 9% had led to a charge, caution or other formal action, and 17% were still open.

Outcomes describe where each investigation stood when the data was published, so recent offences are naturally still open and the closed shares will edge up as cases conclude. Anti-social behaviour is left out entirely: police record no outcomes for it, and counting it would quietly flatter every percentage here. "Formal action" spans charges, summonses, cautions, penalty notices and community resolutions.

Over time

Recorded crime in the Metropolitan Police area peaked at 1,223,176 offences in 2011; the latest 12 months, at 1,130,871, are about 8% below that peak. That works out at roughly 3,098 a day.

Download every month as CSV, including which months each force filed.

Time of year

Recorded crime here peaks in July, about 22% above its quietest month, February.

Crime in context

Officers to record it

Metropolitan Police Service had 31,495 full-time equivalent officers as at 31 March 2026, 3.47 per 1,000 residents, above the England and Wales figure of 2.42. That is 4% above its low point of 30,390 in 2018. Recorded crime partly measures how much policing there is to record it, so a change in officer numbers moves the figures on this page even when nothing else does.

Officer numbers are full-time equivalents from the Home Office police workforce return. Nothing in this section causes anything on this page: a correlation is not a cause.

Stop and search

Metropolitan Police Service carried out 121,505 stop and searches in the 12 months to June 2026, 13.4 per 1,000 people in the force area. 33.4% ended in an arrest or another sanction, against 28.3% across England and Wales. The rest found nothing.

What officers were looking for

Who was searched

Searches set against the number of residents in each group, which is the comparison this question is actually about. A share on its own ("a third of searches were of Asian people") says as much about who lives in the force area as about how it polices.

The two columns are measured differently and that matters. The police record officer-defined ethnicity, meaning what the officer perceived at the time of the search, while the residents column is Census 2021, which is what people say they are. They are close enough to divide and not the same measurement. "Not stated" is searches where no perception was recorded; the census has no such group, so those searches get a count and no rate rather than being folded into a group they may not belong to.

A rate per resident is also not a rate per person present. Searches happen where people are, and a force area's daytime and night-time population is not its resident population, which pushes the rate up for any group over-represented among people travelling into the places searches happen. This table is a starting point for a question, not the answer to one, and the disparity it shows is documented nationally and is not particular to this force.

Home Office stop and search data, published at data.police.uk alongside the crime files. "Found something" counts searches ending in an arrest or another sanction; the rest ended in no further action. Rates use the resident population of the force area, so a force covering a city that fills up by day searches a population larger than the one it is divided by.

Every force compared, and how searches break down by the ethnicity officers recorded, is on stop and search by ethnicity.

When searches happen in the Metropolitan Police area

Of the 373,367 searches recorded in the Metropolitan Police area and timed from July 2023 to June 2026, the busiest hour was 4pm and the quietest 6am. 23% of them fell between 10pm and 4am, against 29% across England and Wales: 6 points less of its searching happens at night.

Metropolitan Police Service England and Wales, for comparison. Both drawn as each one's own share of its searches, so a small area and the country can be put side by side.

Counted on the hour the search happened, in local time. Every search the force filed is counted, because these figures come from the force's own returns rather than from where each search happened. Three years are used rather than one because a single year of one area is too thin to show a shape. Forces filing a date with no time are excluded entirely. Stop and search by time of day sets out the national picture and the method.

Who oversees policing in the Metropolitan Police area

Policing in the Metropolitan Police area is overseen by the Mayor of London, who holds the police and crime commissioner's powers here. Sadiq Khan (Labour) won that election in May 2024. Since that election, recorded crime here is broadly flat (the latest 12 months of data against the 12 months up to the election).

The police force covering an area is answerable to a directly elected commissioner, or to the combined authority mayor where that office has taken the role on. The coloured strip under the trend chart shows these terms against the crime line.

The change column reports what recorded crime did across each term, because it is the first thing a reader wants from this table. It is what happened, not what the officeholder caused: recorded crime moves with reporting rules, how a force records what is reported, national policy and a pandemic, and those effects are far larger than anything one commissioner does in four years. The terms spanning 2020 and 2021 in particular contain the lockdowns, which cut recorded crime everywhere and handed the following term the rebound.

These are election results, not terms of office: commissioners resign, die and are disqualified mid-term, and by-elections between the scheduled polls are listed here only where one has been held. For who holds the office today, the force's own website is the authority. The last column compares recorded crime in the 12 months ending at the handover (or the latest data) with the 12 months ending at the election itself; where a monthly return is missing, the window slides by up to four months to the nearest complete one. A dash means the term ran under a year or no complete window exists near it.

Questions

Is crime in the Metropolitan Police area going up or down?

In the Metropolitan Police area, the latest 12 months are almost level with the 12 before. Over the full series, recorded crime here peaked at 1,223,176 offences in 2011 and the current total is about 8% below it.

What is the most common crime in the Metropolitan Police area?

The most commonly recorded offence in the Metropolitan Police area is violence and sexual offences, with 280,361 offences in the 12 months to June 2026 (25% of all recorded crime). The next most common is anti-social behaviour, with 235,821.

Where does this come from?

Police recorded crime open data, published at data.police.uk by the Home Office under the Open Government Licence. It covers offences the police recorded, which is not the same as offences committed: crimes that are never reported never appear. Rates use ONS mid-year population estimates. How the figures work sets out the calculation, the reporting lag and every known gap.