A common view of buried-explosive seismic sources is that a buried explosion is a pure-P source because it is an expanding volume of high-pressure gas. This assumption leads to a popular conclusion that if shear-mode reflections are observed in explosive-source data, those reflections have to be produced by a downgoing SV wavefield that is generated by P-to-SV conversion at an interface above the depth of the buried explosive. This paper challenges this traditional view that a buried chemical explosive is a pure-P source. In our finite-difference modeling, a buried explosion is a pure-P volume of expanding gas for only that fraction of a microsecond that it takes for its expanding gas to reach the wall of the tube that contains the explosive material. As soon as this accelerating high-pressure gas contacts any surrounding elastic media, a chemical explosion ceases to be a pure-P-source, and shearing is initiated. We chose a finite-difference style of numerical modeling so that we could introduce realistic conditions, such as a physical container for explosive material, a shot hole, small zones of Taylor instabilities in the expanding gases, and local sedimentary interfaces, into numerical calculations. This finite-difference modeling verified that buried explosives not only generate direct-P wavefields, but that they also produce direct-SV wavefields. We believe that our work provides important information that seismic interpreters should consider. Namely, our research indicates that seismic reflection data acquired with buried explosives can be processed to generate direct-SV images in addition to traditional direct-P images. This direct-SV imaging option applies not only to new seismic data acquired with buried-explosives, but also to legacy P data that were acquired with buried-explosive sources many years in the past and now sit dormant in digital seismic-data libraries.