The citizens of Sheffield, Edinburgh, and London are the most likely in the UK to have their mobile phones stolen, a new survey has suggested.

According to card protection company CPP, 11 percent of residents in Sheffield said they had experienced phone theft at some point, slightly ahead of the nine percent and eight percent in the Scottish and UK capitals, respectively. Behind that came Liverpool and Cardiff with 6 percent theft rates each.

A quarter said they had lost phones through pick-pocketing, 20 percent from restaurant and pub tables, while one in ten had been mugged. Five percent blamed colleagues at work.

"With new handsets regularly hitting the market and costing hundreds of pounds, it's no wonder that thieves are looking to target them especially with the increasing amount of personal and sensitive information held on them to access social networking sites and mobile banking services," said CPP's Danny Harrison.

There are obvious problems with subjective crime surveys, starting with extrapolating the statistical likelihood of a crime such as theft happening on the basis of small sample sizes. Adjusting for size, there seems no reason why one city would have more real theft than any other beyond the effect of random sampling.

Sample sizes were small, questioning only 67 people in Edinburgh, 486 in London, and 71 in the hotspot Sheffield out of a total of 2,007 across the UK. What the survey does suggest is that mislaid or stolen phones are a real problem, with many more people losing phones than are ever reported to the police to turn up in official Home Office phone theft figures.

That is partly because most people do not take out insurance and therefore see no reason to report a lost mobile, but also because most mobiles lost are not iPhones or Blackberrys. The average resale value of a recycled mobile - including stolen and recycled phones - is reckoned to be around £25 - a long way from smartphone territory.

The survey follows on from another by CPP in January that claimed Cardiff was the top town for credit card fraud . That was a more plausible finding. Credit cards theft rates can be affected by individual crime rings operating in one town but not in another.