Recent research has provided very strong circumstantial evidence of the existence of neighbourhood effects in voting patterns at recent UK general elections. The usual reason adduced to account for these spatial variations is the neighbourhood effect. This hypothesises that people are influenced in their decisionmaking and behavioural patterns by their neighbours, with interpersonal conversation being the main means of transmitting such influence. Although there is an increasing body of evidence showing the impact of such conversations - that people who talk together, vote together - relatively little of this has grounded the geography of such conversations in the individuals' local neighbourhoods. Those who interact locally should show more evidence of 'neighbourhood-effect-like' patterns than those who do not. To inquire whether this is indeed so, this paper extends recent work on voting patterns in the United Kingdom by investigating the behaviour of individuals with different levels of participation in their local milieux - what we define below as neighbourhood social capital.