This trial studied the behavior of a marker of eosinophilic inflammation, eosinophil cationic protein (ECP), in the peripheral blood of two groups of subjects with seasonal allergic respiratory symptoms (rhinitis and mild bronchial asthma) induced by pollen allergens of Parietaria judaica (P.j.) (one group treated and another untreated with specific immunotherapy [SIT]), to determine what contribution these serial measurements might provide, in comparison with various other tools now available for pollinosis monitoring. In a previously randomized order, we selected 25 patients with monosensitization to P.j. pollen allergens; among them, 12 had started SIT with a P.j. extract in autumn 1993. As a control group, 13 patients were untreated. All patients were studied with various tests at four different times: time I-November 1993; time II-February 1994; time III-end of May 1994; and time IV-September 1994. Blood samples for determination of serum ECP were collected at each time. Methacholine challenge tests were performed at times I and III. A pollen count was also carried out. A statistically significant difference (P < 0.05) was observed in mean ECP levels at times I and III in SIT treated and untreated patients. The interaction between groups and time was not significant. No statistically significant difference was found between PD20 FEV1 values at times I and III in either group. After 1 year of treatment, we did not find any effect of SIT on bronchial hyperresponsiveness or on ECP serum values.