The developmental model of voter turnout posits that the propensity to vote (or abstain) is a habit acquired during the period of a young adult's socialisation to the electoral process. Once acquired, the habit tends to persist and the individual becomes fairly ‘immune’ to the potential influences hypothesised to affect turnout. In particular, Franklin proposes that electoral context will have a strong impact on those who have no or little experience of voting (or abstaining) in democratic elections but virtually no impact on those whose ‘electoral experience’ is more substantial. Tests of the above crucial hypothesis have focused mostly on the heterogeneous effects of electoral competitiveness (election closeness) on turnout. I follow the same route but also point to potential drawbacks of the tests performed so far. At the same time, I propose what I claim is a superior test of experience-conditioned impact of closeness on turnout. My test – utilising data on sixteen Swedish elections conducted between 1956 and 2006 – yields suggestive results supporting Franklin's hypothesis.