Early detection of breast cancer is paramount for improved survival. Mammography, the gold standard for diagnostic breast imaging, has limited spatial resolution and often misses early tumors. In thermoacoustic (TA) imaging an incident microwave pulse is converted to ultrasound via localized thermoelastic expansion and creates an image of the sample's absorption. Single frequency thermoacoustic imaging failed in clinical studies to reliably discriminate malignant from healthy tissue. We hypothesize that limitations in TA technology severely reduced the capability and performance of previous imaging systems. We present initial experimental results using thermoacoustic imaging and spectroscopy for discrimination of fat and muscle tissue, water content and ethanol. This study demonstrates multi-frequency thermoacoustic spectroscopy using both the magnitude and slope of the TA signal between 2.7 and 3.1 GHz. The average slope of the thermoacoustic signal for ethanol was calculated to be -0.45 a.u./GHz, while water had an average TA slope of 0.56 a.u./GHz, clearly differentiating the two regions and consistent with their microwave absorption properties.