Anomalous OH masers were detected towards supernova remnants (SNRs) more than 30 years ago. The satellite line at 1720 MHz was detected in emission, while the main lines (1665 and 1667 MHz) and the other satellite line, at 1612 MHz, occurred only in absorption. Independently, 25 years ago, a theoretical model was proposed whereby 1720 MHz OH masers could be collisionally excited by the passage of a shock through a molecular cloud. For an efficient inversion of the 1720 MHz transition, the gas should have a narrow range of physical properties, namely a kinetic temperature 50 ≤ T
≤ 125 K, a volume density n
2 ∼ 105 cm−3, and a column density of OH gas 1016–1017 cm−2. However, it was not until 1994, with interferometric observations, that the importance of studies of these masers, formed when an SNR is in direct interaction with a molecular cloud, was realised. This discovery triggered a series of surveys to search all known Galactic SNRs. I will review the outcomes of these surveys and discuss the consequences of using 1720 MHz masers as a diagnostic tool for calculating shocked gas conditions, for magnetic field determinations (from Zeeman splitting measurements) and for measuring the properties of SNRs.