This paper aims to assess the relative importance of the principal ‘solvability factors’ in detecting metal crime on the national rail network in England and Wales. It uses British Transport Police data for 4001 metal theft offences committed between 2009 and 2010. Seven main types of metal theft and 12 solvability factors were identified, which include, in order of importance, scrap metal dealer checks, covert police activities, vehicle registration checks, forensic evidence and witnessed offences. The likelihood of detection depended on number of factors present and their effect sizes, with the significant solvability factors accounting for 64 % of variation in detection outcomes. Number of solvability factors available in the case was a significant predictor of solvability with each additional factor boosting detection odds five times. There were notable, countrywide spatial variations in detection rates, some of which reflected solvability differences, while differences in investigative practice are also likely to have played a role. The study provides information for case screening procedures that would improve the cost-effectiveness of metal theft investigations by ensuring resources are directed at the more easily solved cases, not allocating resources to cases which are unlikely to be solved, and by ‘triaging’ lines of enquiry.