Lung surfactant is a complex mixture of phospholipid (PL) and apoproteins that opposes surface tension forces within the lungs, and so prevents alveolar collapse on expiration. The major surfactant component responsible for this reduction in surface tension, dipalmitoyl phosphatidylcholine (PC16:0/16:0), is a solid at body temperature (T,=4 1'C). Consequently, other components of surfactant, including acidic PLs and hydrophobic proteins, have been proposed to facilitate adsorption of PC16:0/16:0 to its site of activity at the air-liquid interface in the lungs [l]. The major acidic PL responsible is thought to be phosphatidylglycerol (PG), which is uniquely enriched in lung surfactant (1 5% total PL) compared with its minor contribution to mammalian cellular PL. Additionally, phosphatidylinositol (PI), present at 7% total PL in human surfactant, has been reported to replace PG in various animal species without altering normal lung function [2]. Limitations of sensitivity inherent in available HPLC methodologies have restricted previous studies of the roles PG and PI composition in surfactant function to animal studies [3,4] and diseases such as alveolar proteinosis (AP) [5] where large sample volumes are available. Consequently, we have developed a sensitive and rapid method for the routine analysis of the acidic phospholipids PG and PI in lung surfactant, based on electrospray ionisation mass spectrometry (ESI-MS). Bronchoalveolar lavage (BAL) (1201111 instilled volume) was collected by fibreoptic bronchoscopy from 12 control adult volunteers who had not exhibited evidence of upper respiratory tract infection within the preceding month. After filtration of mucus and centrifugation at 400 x g x 10 min, the PL concentration of the cell-free BAL fluid supernatant was determined from its phosphate content [6]. Total lipid was extracted from an aliquot of BAL fluid containing 25nmoles of PL using chlorofon/methanol[7] after adding lnmole of dimyristoyl PG (PG14:0/14:0) as an internal standard. The total lipid extract was dried under N2 gas and re-dissolved in 2Opl methano1:chloroform 2:l (v:v) containing 5mM NaOH. An aliquot (5pI) was introduced by rheodyne injection into a methano1:chloroform:water (80: 10: 10, v:v) mobile phase pumped at 1 Opl /min into the electrospray interface of a Quattro II quadrapole mass spectrometer. Spectra were collected at MSl under negative ionisation conditions, and M-IT ions of PG and PI were quantified after correction for the I3C isotope effect. This method was rapid ( ~ 5 midsample) and routinely analysed sample volumes containing about 6 nrnoles acidic PL (0.5-1 ml BAL fluid). Molecular species present at 4 0 0 pmoles could not be resolved. While there was considerable compositional variation between individual subjects, surfactant PG and PI were composed ofthe same three major species, 16:0/18:1, 18:0/18:1 and 18:1/18:1 (+18:0/18:2) (table 1). However, the relative