Introduction: Stigma is a cause of reduced treatment-seeking and delay and has recently been explored as a factor in problem gambling, a population in which 1 in 10 seek treatment (Cunningham, 2005). As stigma has been conceptualized as a process consisting of labels, stereotypes, prejudice, and discrimination, the present three-part study examines stereotypes (study 1), the relative impact of symptoms and labels on desired social distance (study 2), and self-stigma (study 3) in problem gambling. Methods: A qualitative analysis with 152 undergraduates was used to examine problem gambling stereotype content with free response and checklist methodologies (study 1). An online vignette study engaged 801 undergraduates (study 2). Path analysis was used to examine self-stigma and enlisted 159 individuals with gambling problems to complete paper and pencil questionnaires. Results: Problem gambler stereotypes generated included compulsive, irrational, impulsive, risk-taking, antisocial, aggressive, irresponsible, greedy and depressed. Symptom presence accounted for greater variance in social distance than labels and number of symptoms did not impact desired social distance. Labelling led to stigma only in the absence of symptoms. In study 3, the data fit the proposed path model well and self-stigma resulted in reduced self-esteem and increased shame. Shame predicted the negative coping efforts of secrecy and withdrawal. Contrary to expectations, self-stigma predicted increased treatment-seeking. Applying negative stereotypes to problem gamblers as a whole led to reduced treatment-seeking. Additional predictors of treatment-seeking included gambling problem severity, attitudes towards treatment, sex, and income. Conclusions: Implications for problem gambling stigma, for interventions designed for at-risk gamblers and the general public, for problem gambling treatment providers, and for future problem gambling classification are discussed.