Purpose: Sustainability initiatives typically operate for a limited time period, but it is often unclear whether they have lasting effects. The purpose of this paper is to examine a laboratory fume hood campaign, in order to identify factors that might contribute or detract from long-term change persistence. Design/methodology/approach: The University of Toronto Sustainability Office ran a fume hood sash-closing campaign in one building for a four-month period. The campaign had two components: awareness-raising, where the safety and energy benefits of sash-closing were explained through posters, presentations, and a website; and a competition where participants received raffle tickets when their sashes were in compliance during unannounced inspections. Sash heights were recorded six and a half months before the campaign, throughout the campaign itself, and eight months after the campaign. Surveys and focus groups were used to assess participant attitudes and experiences. Findings: The campaign was effective in substantially reducing sash heights while it was in effect. Several months after the campaign, it appeared that sash-closing behaviours had mostly--but not entirely--reverted to pre-campaign levels. Research limitations/implications: This research does not examine differences in responses across users, and it is possible that the campaign was very effective in persuading some individuals. Practical implications: When running short-term behaviour-change campaigns, attention should be given to strategies that enhance or detract persistence of changes. Competitions and prizes could, perhaps counter-intuitively, reduce long-term effectiveness. Originality/value: Persistence is often overlooked in the design of sustainability campaigns. This paper offers important insights into what does not create lasting change and what might. (Contains 6 figures and 2 notes.)