BACKGROUND AND OBJECTIVES: Gambling disorder (GD) is often characterized as an impulsive condition, but results to date have varied substantially by the measure used to assess impulsivity and the modality of the assessment. The purpose of this analysis was to assess the clinical associations between three measures of impulsivity (Eysenck impulsiveness questionnaire [EIQ], Barratt impulsiveness scale [BIS], and stop-signal task [SST]) and GD symptom severity. METHODS: One hundred and fifty-four participants with a current GD diagnosis were assessed on at least one of the three measures of impulsivity and additional clinical measures of gambling severity. Groups within each domain of the measures were divided using a mean split to compare high and low impulsivity (HI and LI) across clinical and demographic variables. RESULTS: Of the included domains, the motor impulsivity domain of the BIS showed multiple clinical differences between the HI and LI groups. The attentional impulsivity domain of the BIS showed limited clinical associations. Surprisingly, HI and LI groups from the SST and EIQ domains did not show any differences in symptom severity. DISCUSSION AND CONCLUSIONS: These findings emphasize the complex nature of impulsivity, particularly as it relates to GD. With disparate results between different measures, it will be important to clarify the specific features assessed by each measure, and their optimal use clinically. SCIENTIFIC SIGNIFICANCE: This analysis suggests that the motor impulsivity feature of the BIS shows the strongest clinical utility for predicting gambling severity. It also emphasizes the importance of sub-typing impulsivity, rather than considering it a single neurocognitive feature.