In this report we discuss the capability of estimating the yield of an explosion from infrasound signals generated by low yield chemical explosions. We used several datasets acquired at distances ranging from near source to the limit of the so called Zone of Silence (300 km). In general, the yield of the explosion is estimated either from the peak amplitude or from the dominant period of the observed signal. Near source observations show that the best way to estimate the yield of explosion is from an integral of the actual overpressure pulse. This method was proven to be superior to either period or peak amplitude. A small dataset of observations at 14 distances, ranging from 20 to 176 km, shows that when similar propagation mechanisms are involved the amplitudes decay in the same way. There is however about one order of magnitude difference among the amplitudes of those arrivals, which current wind corrections fail to explain. A large dataset of observations at fixed distances shows that ranges of yield estimates are about one order magnitude for period based methods and larger for amplitude based methods. The key to improving the estimates may be understanding the propagation.