Research suggests online interventions can have instant impact, however this is yet to be tested with help-seeking adults and in particular those with problem gambling. This study seeks to determine the immediate impact of a single session web-based intervention for problem gambling, and to examine whether sessions evaluated positively by clients are associated with greater improvement. The current study involved 229 participants classified as problem gamblers who agreed to participate after accessing Gambling Help Online between November 2010 and February 2012. Almost half were aged under 35 years of age (45 %), male (57 %) as well as first time treatment seekers (62 %). Participants completed measures of readiness to change and distress both prior to and post-counselling. Following the provision of a single-session of counselling, participants completed ratings of the character of the session (i.e., degree of depth and smoothness) post-counselling. A significant increase in confidence to resist and urge to gamble and a significant decrease in distress (moderate effect size; d = .56 and .63 respectively) was observed after receiving online counselling. A hierarchical regression indicated the character of the session was a significant predictor of change in confidence, however only the sub-scale smoothness was a significant predictor of change in distress. This was the case even after controlling for pre-session distress, session word count and client characteristics (gender, age, preferred gambling activity, preferred mode of gambling, gambling severity, and preferred mode of help-seeking). These findings suggest that single session web-based counselling for problem gambling can have immediate benefits, although further research is required to examine the impact on longer-term outcomes.