Background: Heroin use disorder (HUD) remains one of the gravest public health issues in China. Methadone maintenance treatment (MMT) and protracted abstinence (PA) are the most commonly used treatments for HUD. Although both treatment approaches can alleviate heroin cravings, a previous study found that MMT patients had stronger cue-induced brain activation than patients undergoing PA; however, the changes associated with long-term treatment are unclear. Methods: Male patients with HUD who had been undergoing either PA (n=24) or MTT (n=21) for approximately 12 months, together with 20 demographically matched healthy controls, completed an event-related functional magnetic resonance imaging (fMRI) task. The subjective craving for heroin was evaluated using a visual analog scale. Results: Compared to the healthy controls, the MMT and PA groups demonstrated significantly higher brain activation in the left pallidum, middle occipital gyrus, postcentral gyrus, anterior cingulate cortex, middle cingulate cortex, inferior parietal lobule, superior parietal lobule, amygdala, hippocampus, right inferior temporal gyrus, inferior frontal gyrus triangularis, and caudate during exposure to heroin-related cues. Compared to those undergoing PA, patients in the MMT group demonstrated significantly higher brain activation in all of these regions. Except for the left inferior parietal lobule and left superior parietal lobule, there were no statistically significant differences between the PA and healthy control groups. The MMT patients showed significantly higher subjective cravings before and after exposure to heroin cues than the PA group, but there was no significant difference in the change in subjective cravings between the 2 groups. Conclusions: The results suggested that although the HUD patients receiving long-term MMT complied with the treatment, they still had higher subjective cravings and cue-induced brain activation than those undergoing PA. Therefore, long-term PA appears to be more beneficial than MMT in reducing the salience value of drug cues in patients with HUD.