Sinusoidal signals with time-varying amplitude show up in many signal processing applications, such as in radar applications when the target is spatially dispersed. Amplitude modulation results in an enlargement of the signal subspace, i.e. the signal subspace corresponding to one amplitude-modulated sinusoid is no longer spanned by a single vector. The authors propose two subspace-based techniques to estimate the centre frequency of a sinusoidal signal with random ARMA amplitude. The proposed techniques are obtained by using the principles behind the ESPRIT and MODE approaches of frequency estimation and by appropriate adaptations and simplifications following careful considerations of the specifics of the problem considered. Numerical simulations illustrate the good performance of the methods. A robustified version of the proposed methods is described and succesfully applied to real radar data.