Very small organisms get the oxygen they need for life by diffusion across their body surface, but larger organisms need specialized respiratory organs to do the same job. In insects, the main respiratory organs are the tracheae,an internal tubular network that develops from clusters of ectodermal cells on either side of the thoracic and abdominal embryonic segments. Two papers in this issue of Development provide new information about the evolution and development of tracheae. On p. 785, Franch-Marro and colleagues raise the possibility that an evolutionary relationship exists between insect tracheae and the external respiratory gills of crustaceans. These gills are associated with appendages, and Franch-Marro et al. show that Drosophila tracheal placodes arise next to leg primordia in the thoracic segments and next to cryptic leg primordia in the abdominal segments;the different fates of the tracheal placode and leg primordia are controlled by Wingless signalling. The researchers also report that homologues of tracheal-inducing genes are expressed in the developing gills of crustaceans. Based on these results, they propose that the ancestors to arthropods had areas on the surface of their body that were specialized for gas exchange,which evolved into crustacean gills and insect tracheae. On p. 957, Matusek and co-workers reveal new details about the development of the Drosophilatracheal system by reporting that the formin DAAM (Dishevelled-associated activator of morphogenesis) regulates the tracheal cuticle pattern. The tracheal cuticle resembles a corrugated vacuum-cleaner hose. This structure gives the tracheae rigidity but allows them to bend as the insect moves. The researchers report that in the absence of DAAM, an array of actin cables beneath the apical surface of the tracheal cells fails to form properly- formins are key regulators of the cytoskeleton - and consequently the pattern of ridges (taenidial folds) in the tracheal cuticle is disrupted and the tracheal tubes collapse. Other results indicate that DAAMactivity is regulated by RhoA and that DAAM works with the non-receptor tyrosine kinases Src42A and Tec29 to organise the actin cytoskeleton and thus determine the cuticle pattern of Drosophila tracheae.