The dynamics of fixation and release of NH4+ in soils were studied using tracer N under field and laboratory conditions. Field data showed that release of fixed NH4+ was relatively slow after an initial moderately fast release. Forty months of field weathering of Bainsville soil left 3.48 kg 15N/ha in the 75-cm profile of the 13.5 kg 15N/ha applied and most (76%) of this recovered 15N was fixed NH4+–N. The relative quantitative importance of recently fixed NH4+ in the various particle size fractions was not in the same order as the native fixed NH4+. The fine silt fraction (2–5 μm) fixed a larger amount (whole soil basis) than the fine clay fraction (< 0.2 μm). The coarse clay fraction (0.2–2 μm) fixed the most NH4+ added as well as being the fraction containing the most native fixed NH4+. Sand size fractions were shown to contain native fixed NH4+ and were capable of fixing a small amount of added NH4+. Measurements of recently fixed NH4+–N in various particle sizes covering four time intervals (up to 40 mo) of field weathering under fallow showed that the 0.2 to 2-μm fraction was quantitatively most important, the < 0.2-μm fraction most readily released recently fixed NH4+ and the 2 to 5-μm fraction was the most stable with respect to recently fixed NH4+. Plant growth affected the relative stability of the recently fixed NH4+ in the various particle size fractions. Laboratory studies of three Brookston soil samples revealed higher estimates of fixation of NH4+ by direct tracer methods than by indirect extraction methods. This discrepancy was assumed to have been caused by some exchange of added NH4+ with native fixed NH4+ and by some NH4+ being fixed on other sites. The relationship between mineralogical and related analyses to fixation of NH4+ was discussed.