The purpose of this study was to examine the relationships between types of gambling and disordered gambling, with and without controlling for gambling involvement (i.e. the number of types of games with which respondents were involved during the past 12 months).We completed a secondary data analysis of the 2007 British Gambling Prevalence Survey (BGPS), which collected data in England, Scotland and Wales between September 2006 and March 2007. The sample included 9003 residents, aged 16 or older, recruited from 10 144 randomly selected addresses. 5832 households contributed at least one participant. Post-facto weighting to produce a nationally representative sample yielded 8968 observations. The BGPS included four primary types of measures: participation in gambling (during the past 12 months and during the past 7 days), disordered gambling assessments, attitudes toward gambling and descriptive information.Statistically controlling for gambling involvement substantially reduced or eliminated all statistically significant relationships between types of gambling and disordered gambling.Gambling involvement is an important predictor of disordered gambling status. Our analysis indicates that greater gambling involvement better characterizes disordered gambling than does any specific type of gambling.