We investigated the frequency and prognostic significance of qualitative olfactory dysfunction (parosmia, phantosmia) in a retrospective patient based study. A total of 392 patients with impairment of olfaction were tested at least two times for their olfactory function using the "Sniffin' Sticks". The mean interval between the first and the last test was 11 months. At the first visit 34% of all patients reported parosmia. Parosmia was most frequent in patients with postinfectious olfactory loss (56%), and less frequent in idiopathic, posttraumatic, sinunasal disease with frequencies of 10, 14, and 28%, respectively. In contrast, only 12% of all patients had phantosmias, with no significant differences between the patient groups. Improvement of olfactory function was found in 23% of all patients (n = 90). Pre-existing parosmia or phantosmia had no significant effect on recovery rate. Regarding qualitative olfactory dysfunction, 29% of those patients reporting parosmia reported relief of this symptom after an average of 12 months, whereas 53% of phantosmic patients lost phantosmia during the observation period. Although it has been suggested that olfactory distortion s could be regarded as an indicator of early recovery of decreased olfactory sensitivity, the current data indicate that occurrence of parosmia or phantosmia has little prognostic value. Phantosmia disappears at a faster rate than parosmia. These insights into qualitative olfactory dysfunction are regarded to be significant in the counseling of patients with olfactory loss.