Abrupt thrombosis is a form of thrombosis that occurs unexpectedly and without being preceded by hemodialysis fistula (AVF) dysfunction during dialysis. We found that AVFs with a history of abrupt thrombosis (abtAVF) appeared to have more episodes of thrombosis and required more frequent interventions than those without such history. Therefore, we sought to characterize the abtAVFs and examined our follow-up protocols to determine which one is optimal. We performed a retrospective cohort study using routinely collected data. The thrombosis rate, AVF loss rate, thrombosis-free primary patency, and secondary patency were calculated. Additionally, the restenosis rates of the AVFs under the follow-up protocol/sub-protocols and the abtAVFs were determined. The thrombosis rate, procedure rate, AVF loss rate, thrombosis-free primary patency, and secondary patency of the abtAVFs were 0.237/pt-yr, 2.702/pt-yr, 0.027/pt-yr, 78.3%, and 96.0%, respectively. The restenosis rate for AVFs in the abtAVF group and the angiographic follow-up sub-protocol were similar. However, the abtAVF group had a significantly higher thrombosis rate and AVF loss rate than AVFs without a history of abrupt thrombosis (n-abtAVF). The lowest thrombosis rate was observed for n-abtAVFs, followed up periodically under the outpatient or angiographic sub-protocols. AVFs with a history of abrupt thrombosis had a high restenosis rate, and periodic angiographic follow-up with a mean interval of 3 months was presumed appropriate. For selected populations, such as salvage-challenging AVFs, periodic outpatient or angiographic follow-up was mandatory to extend their usable lives for hemodialysis.