Maximum acceptable preextraction analytical holding times (MHTs) were estimated by spiking aqueous solutions of two nitramines, octahydro-1,3,5,7-tetranitro-l,3,5,7-tetrazocine (HMX) and hexahydro-l,3,5-trinitro-l,3,5-triazine (RDX) and three nitroaromatics, 1,3,5-trinitrobenzene (TNB), 2,4,6-trinitrotoluene (TNT), and 2,4-dinitrotoluene (2,4-DNT), into three soils. Nitramines were stable over an 8-week test period at all storage temperatures, but nitroaromatics, which were reasonably stable when frozen, degraded rapidly at room temperature and more slowly under refrigeration. In contrast, both nitroaromatics and nitramines were quite stable under refrigeration for four field-contaminated soils. When three of these field-contaminated soils were subsequently fortified with TNT and TNB, rapid degradation under refrigeration was again observed for the added nitroaromatics. The rates of degradation were quite different in the three soils, further emphasizing the uncertainties in MHT estimates derived from fortified soils. We conclude that fortified soils can produce very different estimates of MHTs, recoveries, treatability, etc., compared to field-contaminated soils even when spiking is done in aqueous media.