Biological ammonia oxidation had long been thought to be mediated solely by discrete clades of - and -proteobacteria (ammonia-oxidizing bacteria; AOB). However, ammonia-oxidizing Crenarchaeota (ammonia-oxidizing archaea; AOA) have recently been identified and proposed to be the dominant agents of ammonia oxidation in soils. Nevertheless, the dynamics of AOB versus AOA, and their relative contribution to soil ammonia oxidation and ecosystem functioning on stress and environmental perturbation, remain unknown. Using a 3-year longitudinal field study and the amoA gene as a molecular marker, we demonstrate that AOB, but not AOA, mediate recovery of nitrification after zinc (Zn) contamination. Pristine soils showed approximately equal amoA gene copy numbers and transcript levels for AOB and AOA. At an intermediate Zn dose (33.7 mmol Zn per kg), ammonia oxidation was completely inhibited, and the numbers of AOB and AOA amoA gene copies and gene transcripts were reduced. After 2 years, ammonia oxidation in the field soils was fully restored to preexposure levels, and this restoration of function was concomitant with an increase of AOB amoA gene copy and gene transcript numbers. Analysis of the restored community revealed domination by a phylogenetically distinct Zn-tolerant Nitrosospira sp. community. In contrast, the numbers of AOA amoA gene copies and gene transcripts remained 3- and 104-fold lower than recovered AOB values, respectively. Thus, although recent findings have emphasized a dominant role of archaea in soil-borne ammonia oxidation, we demonstrate that a phylogenetic shift within the AOB community drives recovery of nitrification from Zn contamination in this soil.