Desmosine, the intermolecular and intramolecular cross link between the chains of elastin polypeptide, may be useful as a marker of a lung injury in adult respiratory distress syndrome (ARDS). A radioimmunoassay for rabbit antibody developed against desmosine, conjugated to bovine serum albumin, can detect as little as 100 pg of desmosine in plasma or urine. Desmosine is not metabolically absorbed, reused, or catabolized by the body, but rather eliminated unchanged in the urine as low molecular weight peptides. The lung is relatively rich in elastin, and we reasoned that a timed collection could be used as an index of elastin degradation in vivo. A 2-h collection of urine for desmosine assay was obtained at the time of Swan-Ganz catheter insertion in 41 consecutive patients. On the basis of clinical and initial Swan-Ganz catheter data, the patients were assigned to one of three groups: an ARDS group (n = 12); a cardiogenic pulmonary edema (CPE) group (n = 12); and a critically ill, nonpulmonary edema group (NPE, n = 17). The mean urine desmosine concentration (mg/L) for the ARDS group (0.728 +/- 0.22 SE) differed from the CPE group (0.149 +/- 0.07; p less than 0.001). The total excretion (microgram/2 h) was 64.95 +/- 24.7 in the ARDS group and 24.71 +/- 11.7 in the CPE group (p less than 0.05). Urine desmosine concentration/serum creatinine index for the ARDS group (0.78 +/- 0.28) was greater than in the CPE group (0.07 +/- 0.04; p = 0.019). Desmosine excretion was increased in the NPE group compared with CPE and ARDS groups, possibly reflecting heterogeneity in this group. In the differentiation of ARDS from CPE, we conclude that substantial increases in urinary desmosine excretion favor a diagnosis of ARDS.