Abstract: Crowdfunding, and research which studies it, has exploded in the last decade. Every year, we learn more about the different types of crowdfunders, including ‘serial crowdfunders’ – people who go on to launch several campaigns in succession. Nearly one-third all of funds on the leading crowdfunding platform Kickstarter are undertaken by serial crowdfunders. We still know relatively little about whether success at raising funds in one crowdfunding campaign improves serial crowdfunders’ success prospects in subsequent campaigns. Answering this research question promises to enhance our understanding of the financing of a sequence of ventures by the same entrepreneur. Prior work has explored multi-stage lending of venture capital-backed projects, and relationship lending between banks and borrowers, but analysis of the financing of multiple distinct projects by serial entrepreneurs scarce. By exploring sequences of fundraising success among serial crowdfunders, we are able to discover whether or not past success leads to future success – a question of keen interest to entrepreneurs and their backers on crowdfunding platforms. To investigate this topic, we gather data on serial crowdfunders from the world’s largest reward-based crowdfunding platform, Kickstarter. Our dataset is comprised of 29,788 serial crowdfunders who launched 75,654 campaigns since Kickstarter’s inception in April, 2009 and up to November, 2016. In total they raised $859 million, which represents 38.5% of the funds raised on Kickstarter during the same period. Our analysis reveals for the first time that amounts raised by serial crowdfunders exhibit a cyclical pattern over time, with above-average funding success being followed by below-average success, which is then followed by above-average success, and so on. Our empirical investigations suggest that effort-based dynamic reputation management can explain this finding. That is, following an over-performing crowdfunding campaign, entrepreneurs can afford to supply somewhat less costly effort to the next campaign without unduly damaging their reputation with backers, resulting in underperformance; but this underperformance stimulates entrepreneurs to exert higher levels of effort next time to restore their reputation, leading to over-performance. Our findings carry several implications for scholarship and practice. For example, our work highlights the importance of crowdfunders’ reputations and reputation management. Thus, if an entrepreneur anticipates wanting to launch a future project about which they are especially passionate, our results recommend they exert higher effort than they might otherwise supply to that project’s predecessor. This would proactively build a positive reputation which would increase the future project’s chances of getting funded. Additionally, our results suggest that serial crowdfunders may want to develop industry-specific knowledge by launching campaigns in the same industry, since changing industry seems to adversely affect their subsequent fundraising performance. We also highlight the role of timely launches of new campaigns, since waiting for longer periods of time between campaigns seems to adversely affect funding outcomes as well. Finally, our results also suggest that crowdfunding platform operators might want to provide additional guidance to crowdfunders beyond their initial campaigns. Instructions helping crowdfunders to maintain their fundraising performance in subsequent campaigns, or how to reverse unfavorable funding outcomes, could also be of considerable practical value.