<b>Background:</b> On average, asthma patients have lower lung function (LF) than healthy individuals and their lung function (FEV1) decline may be accelerated over time. <b>Methods:</b> In a cross-sectional analysis on severe asthmatics enrolled in the Belgian Severe Asthma Registry with at least 2 visits minimum 12 months apart, we compared characteristics of patients with and without LF decline (loss of post-BD FEV1 (% predicted)/year greater than zero) over time. Multiple imputation chained equation was applied to study the factors independently associated with FEV1 decline. <b>Results:</b> In the overall population (n=322), annual FEV1 decline was -0.24 (-5.16; 3.01) %pred/yr over a period of 22 (12-41) months. In the LF declining group (n=145, 45%), patients were slightly older and their FeNO levels were lower, whereas post-BD FEV1, post-BD FVC, and post-BD FEV1/FVC were lower and asthma less controlled in non-decliners. Twenty four percent of non-decliners were started on anti-IL5 during the study compared to 8% in decliners. The multivariable linear regression model data suggested that for every additional unit of ACQ score at baseline, there was a decrease of 3 % in decline (95% CI, -4.69 to -1.17; p=0.001) and decline was reduced by 15 % (95% CI, -20.17 to -9.29; p<0.0001) in patients in whom an anti-IL5 was started during follow-up compared to patients without anti-IL5 therapy (Fig.1), independently from other changes (in exacerbations, smoking status, ICS or OCS dose, or add-on anti-IgE). <b>Conclusion:</b> Targeting severe eosinophilic asthma with anti-IL5 add-on therapy could prevent LF decline. This should be confirmed in future prospective controlled studies.