Acoustic emission (AE) techniques have obvious attractions for structural health monitoring (SHM) due to their extreme sensitivity and low sensor density requirement. A factor preventing the adoption of AE monitoring techniques in certain industrial sectors is the lack of a quantitative deterministic model of the AE process. In this paper, the development of a modular AE model is described that can be used to predict the received time-domain waveform at a sensor as a result of an AE event elsewhere in the structure. The model is based around guided waves since this is how AE signals propagate in many structures of interest. Separate modules within the model describe (a) the radiation pattern of guided wave modes at the source, (b) the propagation and attenuation of guided waves through the structure, (c) the interaction of guided waves with structural features and (d) the detection of guided waves with a transducer of finite spatial aperture and frequency response. The model is implemented in the frequency domain with each element formulated as a transfer function. Analytic solutions are used where possible; however, by virtue of its modular architecture it is straightforward to include numerical data obtained either experimentally or through finite element analysis (FEA) at any stage in the model. The paper will also show how the model can used, for example, to produce probability of detection (POD) data for an AE testing configuration.