Long-term treatment with nucleos(t)ide analogs (NAs) is the current first line therapy for patients with chronic hepatitis B (CHB), recommended by most of the current guidelines. NAs prevent disease progression, liver failure, decrease the risk of hepatocellular carcinoma (HCC), and have favorable safety profiles. However, low rates of on-therapy functional cure (hepatitis B surface antigen [HBsAg] loss), which is regarded as the optimal end point, prevent many patients from stopping NA therapy with the need for a lifelong treatment. The higher likelihood of HBsAg loss associated with stopping as compared to continuing NAs has got a lot of attention recently. Recommendations regarding endpoints allowing for safely stopping NA therapy differ between international guidelines. Whereas in HBeAg-positive patients, HBeAg seroconversion with at least one year of consolidation therapy is an acceptable endpoint of treatment, the recommendations for HBeAg-negative ones differ. Some guidelines propose ≥3 years of HBV DNA undetectability to stop NA while others regard HBsAg loss as the only acceptable endpoint. Stopping NA can lead to substantial rates of virologic relapses and consequent ALT flares in some cases. Moreover, no reliable predictor(s) of post-NA relapses have been identified so far. Quantitative HBsAg is becoming an increasingly promising marker to predict safe NA cessation. On the other hand, investigating the role of the immune system in mediating sustained virologic responses after NA withdrawal is needed to suggest immunological biomarkers to safely stop NA. In this article, we will review relevant literature regarding NA stopping strategy and discuss promising viral and immunological biomarkers to predict antiviral responses and thus to help identify patients who are more likely to achieve HBsAg seroclearance.