OBJECTIVES: Uncertainty exists about the most meaningful measure of health utilities for treatment comparisons, especially for conditions with lower severity. This study explores these issues for patients with allergic rhinitis (AR), which is a bothersome and costly condition, but is not highly morbid. Inclusion criteria for trials may select only patients with mild or moderate symptom severity; furthermore, these patients often have access to symptomatic medications. Consequently, discrimination of treatment benefit may be difficult. METHODS: The study reviews and compares the measurement properties of the EQ-5D and SF-6D, with particular focus on their ability to discriminate at lower levels of severity, a circumstance of AR. Data from several studies of patients with AR demonstrate that a subset of patients have severity of symptoms that can be meaningfully captured with the EQ-5D (i.e., they are not at the ceiling), while a large subset of patients have lower severity and may be better assessed by the SF-6D.RESULTS: Several published studies of patients with AR and data from this research team indicate that the EQ-5D values are often at or near the ceiling (mean values = 0.94), reflecting low symptom severity. Consequently, the ability to discriminate and make meaningful resource allocations may be compromised. Data from our naturalistic study found that half of the patients with AR reported low symptom severity (mean score of 1.3 on a scale from 0 to 3), suggesting that EQ-5D values would be close to 1.0. CONCLUSIONS: For conditions with lower overall severity and morbidity, such as AR, using more than one utility measure may be informative. The EQ-5D would provide more international comparability, since it is the more generally accepted utility measure. However, inclusion of another utility measure that can capture the lower end of symptom severity may be a useful strategy for patients with AR.