PURPOSE Currently, little evidence supports the safety of suspending vascular endothelial growth factor (VEGF) inhibitors for neovascular age-related macular degeneration (nAMD). We assessed the outcomes of eyes in which this seems to have been attempted. DESIGN Observational study from a prospectively designed database. PARTICIPANTS Eyes enrolled in the Fight Retinal Blindness! registry of nAMD treatment outcomes were considered to have suspended treatment if they had a 3-month or longer documented period of inactivity of the choroidal neovascular lesion with no further treatments unless the lesion re-activated. METHODS Time and proportion to re-activation of the lesion were analyzed using Kaplan-Meier survival curves. Visual outcomes after treatment suspension were assessed with paired t tests. MAIN OUTCOME MEASURES The proportion of eyes resuming treatment because of lesion re-activation, change in visual acuity (VA) at time of re-activation, and recovery of vision 12 months later. RESULTS We identified 434 eyes in which treatment was suspended and that were tracked for at least 12 months thereafter. The estimated percentage of eyes re-activating in the first year after treatment suspension was 41%, increasing to 79% by the fifth year. The median time to re-activation was 504 days. The 275 eyes whose lesion was observed to re-activate lost a mean of 4.2 letters (95% confidence interval [CI], -5.6 to -2.8 letters; P < 0.001) from the last injection to the time of re-activation; 206 eyes resumed treatment for at least 12 months after re-activation and recovered a mean of +1.2 letters (95% CI, -0.4 to 2.7 letters; P = 0.133), resulting in a net loss of 3.3 letters (95% CI, 2.3-5.1 letters; P < 0.001) compared with VA at treatment suspension. Lower VA at the time of suspension and longer duration of treatment were associated with reduced risk of re-activation. Median time to re-activation was substantially greater when eyes had been treated for at least 3 years. CONCLUSIONS Fewer than half of the eyes in which treatment was suspended re-activated in the first year, but most re-activated by the fifth year. Caution should be exercised to avoid suspending treatment prematurely. Further research is warranted to identify the eyes in which treatment may be suspended safely.