Most tissue characterization methods use RF data originating from backscattering processes in the insonified tissue. Characterization of the tissue of a thin layer (e.g., an artery wall) poses problems because of the small amount of data available. The tissue layer may be so thin that there are no RF data that are not corrupted by the reflections from nearby interfaces. An alternative approach may be to compare the spectra of the reflections at the boundaries of the layer for the characterization of the tissue in between. To avoid problems that arise from the use of short (broadband) pulses the method proposed here uses long and almost monochromatic pulses. The frequency dependence of the acoustic parameters can be assessed by using different central emission frequencies. A deconvolution algorithm is used to separate overlapping reflections from closely spaced tissue interfaces. Wiener inverse filtering together with correlation of deconvolved signals results in an accurate determination of the position of specular reflectors. The accuracy of the peak values detected in the envelope signal is not sufficient for quantitative determination of tissue parameters, but still much better than results from the analysis of short waveforms.