The objective of this paper was to compare distortion of the tie wings and bases of metallic orthodontic brackets following clinical use and after debond by either of two methods, and took the form of a prospective random control trial. Five-hundred-and-seven brackets were debonded using either bracket removing pliers or a lift off debonding instrument (LODI). By a system of random allocation contralateral opposing quadrants were debonded with a 0.019 x 0.025-inch archwire either in place or removed. After debond brackets were tested for slot closure by the fit of rectangular test wires from 0.016 x 0.022 to 0.021 x 0.025 inch in size. The LODI produced few slot closures sufficient to affect the fit of all but the largest test wire. Bracket removing pliers used after removal of the archwire produced significantly greater numbers of distorted brackets in response to testing with all but the largest wire. With the 0.021 x 0.025 inch wire in place the presence or absence of the archwire at the time of debond made no difference to the number of slot closures. Ten per cent of the brackets debonded using bracket removing pliers had distorted bases, no base damage was produced by the LODI. The use of bracket removing pliers at debond caused significantly more slot closures than use of the LODI. When bracket removing pliers are used the archwire should be left in place at the time of debond since this reduces the number of distortions.