The purpose of this article is to draw attention to two problems encountered with modern aircraft: the difficulties in making adhesive and paint adhere to composite substrates and the lack of any after-the-fact inspection that can prove that there will not be any interfacial failures at some time during the service life. It is also observed that the response to paint peeling off is more rapid and thorough than to a discovery of separations between internal components that were once believed to have been bonded together. Because there is so much similarity between the processes of making paint and adhesive adhere, it is suggested that some of the efforts to improve adhesion of the paint might also help improve the processes for making adhesives stick. The article focuses on a series of anecdotes about problems and their resolutions, with the hope that the solutions might help others solve or avoid future such problems. It is pointed out that the cost of improving the adhesion of both paint and adhesive has always been insignificant in comparison with the sometimes enormous costs incurred as a result of fleet-wide occurrences of what were perceived to be bond "failures but which should more properly be characterized as initially undetected nonbonds. A critical issue is the acknowledged absence of any nondestructive inspection capable of distinguishing between bonds that will "fail'' in service and those that will not. Experience has shown that none of the apparent interfacial failures to date have occurred on grit-blasted surfaces. Equally, it must be conceded that not all of the bonded composite structures made using peel-ply surfaces can be expected to fail, even though those associated with released peel plies or prebond moisture probably will, because these conditions have been associated with so many of the past failures. The distinction between interfacial failures and impact damage to properly bonded structures is that the former can extend throughout the entire structure, whereas the broken fibers and interlaminar matrix failures associated with the latter will not extend far beyond the impact area. This is one reason why it is so important to use only surface preparations that ensure the absence of interfacial failures. It is also noted that there is no counterpart, for the bonding of composite structures, of the peel-type test that was so instrumental in solving the equivalent bonding problem that was widespread in bonded metal structures some 30 years ago. It is recommended that there should be, because the use of only shear-load tests has been found to be insufficient to ensure bond durability for both metallic and composite structures.