Thermoacoustic imaging (TAI) is an emerging promising hybrid modality for biomedical imaging. High spatial resolution of TAI is conventionally accomplished by utilizing a short exciting microwave pulse, but squeezing the microwave pulse simultaneously degrades the signal-to-noise ratio (SNR) of the thermoacoustic image. There is usually a tradeoff between resolution and SNR. Compressive sensing (CS), a novel algorithm for signal recovery, is introduced into TAI to obtain images whose resolutions are largely independent of the microwave pulse width. In this work, the algorithm based on CS is applied to reconstruct images for different widths of rectangular microwave pulse varying from 0.5 to 2-μs. It is shown that the strength of the generated acoustic signals can be enhanced without compromising the resolution of the reconstructed images as the microwave pulse width increases.