The effect of anaesthetic and other vapours on lung surfactant has been investigated by four methods. Bubbles obtained from the lung were exposed in hanging drops to various concentrations of the vapours of ether, chloroform, and halothane. At 36.5 °C the bubbles lost their stability when the vapour had more than about 40% of its saturation concentration. Such a concentration is much greater than any used in practice. Organic vapours usually destabilize the bubbles when saturated. When a collapsed lung was reinflated with air saturated with certain rather insoluble vapours, it tended to collapse again; with other vapours it remained open. Mice killed with massive doses of insoluble vapours (e.g. pentane) showed lung collapse; those killed with more soluble vapours (e.g. ether) did not. Prolonged exposure of collapsed lung to saturated anaesthetic vapour did not prevent it from forming a lining film, but the film was somewhat abnormal. These findings are discussed in relation to anaesthesia, and it is concluded that collapse of the lining film would not occur in normal practice and would be a hazard only in the event of a gross overdose of halothane. It is concluded that acute massive collapse of the lung, as reported in the literature, is unconnected with the action of anaesthetic on the lining film. The means adopted to produce partially saturated vapours are discussed in relation to oil-gas partition coefficients. A method of testing bubbles for “clicking”, using perfluoromethylcyclohexane vapour, is described.