Eosinophilic infiltration is a prominent component of the inflammation characteristic of asthma. Measurement of eosinophils and eosinophil cationic protein (ECP) in induced sputum is a noninvasive indicator of changes in airways inflammation. Recently, Schmekel and colleagues1 reported that salivary ECP was elevated in subjects with asthma and was correlated with disease activity. The measurement of salivary ECP as a surrogate for sputum ECP is attractive because it can be done rapidly, requires less cooperation on the part of the subject, and is devoid of the potential adverse effects of sputum induction. We examined the correlation of sputum ECP and salivary ECP in 26 paired samples from 14 nonsmoking asthmatic subjects not receiving corticosteroids (12 subjects had specimens collected 4 weeks apart). Asthma severity ranged from mild intermittent to moderate persistent. After mouth rinsing, 3 mL of saliva was collected by repetitive expectoration. Sputum was then induced by inhalation of 3% saline solution and collected over 20 min. Specimens were collected on ice and processed within 60 min. ECP in sputum and saliva was measured by radio-immunoassay (Kabi Pharmacia Diagnostics; Uppsala, Sweden). Mean (± SD) values for sputum and salivary ECP were 88 ± 141 μg/L and 43 ± 20 μg/L, respectively. Following acid acetate extraction, as reported by Schmekel and colleagues,1 there was an increase in salivary ECP to 201 ± 189, likely reflecting lysis of cells. There was no correlation between sputum and saliva ECP (r2 = 0.00). After acetate extraction, the correlation was better (r2 = 0.26, p = .029) though not strong. Hence, while salivary ECP may be elevated in subjects with asthma compared to normal subjects, it is not useful as a surrogate marker for sputum ECP.