Impulsive interference is an elusive phenomenon which cast a shadow on the initial success of Digital Terrestrial Television in the UK at the time of launch of ITV Digital. The unavailability of suitable test methods for impulsive noise has also hindered the search for new measures to counter its effect. Until now, gated Gaussian noise had been widely recognised as the only calibrated tool available to measure impulsive interference performance. From October 2001 to October 2002 a working group within the DTG led by BBC R&D carried out a series of theoretical and practical studies to devise a representative set of test waveforms for impulsive interference. A statistical parametric model for impulsive interference based on randomly pulsed bursts of Gaussian noise was fitted to a large set of captures of real impulsive noise. This model was extensively used in laboratory measurements and its applicability was validated. The result of all this work was the proposal of a group of new 'gated-squared' Gaussian noise tests. Some recommendations for test methods and measurement equipment were also made. As a by-product of this measurement campaign came the realisation that the impulsive noise performance of a DTT receiver equipped with no specific countermeasures can be determined from the effective duration of the burst of noise.