Background: Immunoglobulin E (IgE) is the treatment target of omalizumab, a monoclonal antibody indicated in the treatment of severe allergic asthma. Long-term variability of serum total IgE (sIgE) in asthmatics remains poorly documented. Methods: In this prospective study, sIgE levels were measured over 1 year at 7 time points in 41 severe asthmatics treated with high-dose of inhaled corticosteroids and long-acting β agonists. 33 patients were atopic based on at least one positive RAST to common aeroallergens. Patients were divided into three groups according to their baseline sIgE level: low (< 76 IU/mL; n = 10), intermediate (76-700 IU/mL; n = 20) or high (> 700 IU/mL; n = 11). Patients also completed the six-item Juniper Asthma Control Questionnaire (ACQ). The sIgE variability and factors predictive for this variability were studied, as well as ACQ outcomes. Results: The variation in sIgE level was mostly the consequence of between patient-variability, which represented 96%, 71% and 96% of the total variability in the low, intermediate and high sIgE subgroups, respectively. The residual within-patient variability was therefore limited. In 10/41 patients, sIgE levels increased or decreased, for at least one visit, beyond the predefined range of the subgroups to which they were assigned (< 76 IU/mL; 76-700 IU/mL; > 700 IU/mL). There was a significant but weak correlation between sIgE and ACQ score over all time points (r = 0.15, p = 0.02), but sIgE failed to associate with severe exacerbation. sIgE decreased by 3% with any additional year of age for the whole group (p = 0.01) and increased by 5% per one unit of allergen exposure score in atopic patients (p = 0.002). Conclusion: In severe asthmatics, limited within-patient variability of sIgE levels was observed over 1 year as opposed to marked between-subject variability. sIgE decreases with age. Variation in sIgE weakly associates with asthma control but not with exacerbation.