Estimates of the dry deposition of ammonia (NH3) gas in a field fumigation experiment on an ombrotrophic bog have been made using the inferential technique, with measured wind speed at 2m, and air concentrations at two heights above the vegetation. The parameters for a concentration-dependent surface resistance term have been derived from flux measurements over the same vegetation in a chamber study, separating stomatal from non-stomatal resistances.Annual NH3-N deposition in each of the 4 years 2003–2006 was estimated to increase from 3.0±0.2kgNha−1y−1 in ambient air, with an NH3 concentration at 0.5m above the canopy of 0.7μgm−3, to 50–70kgNha−1y−1 where annual average air concentrations were 70–90μgm−3 and concentrations during fumigation were up to 1600μgm−3. The equivalent deposition velocities (at z=0.5m) were 0.016ms−1 in ambient air and 0.003ms−1 at 100μgm−3. The differences between annual deposition estimates made from independent air concentration data at 0.1m and 0.5m above the canopy were small for distances more than 10m from the source, after vertical mixing was complete. Over 4 years (2003–2006) and at eight sampling points more than 10m from the NH3 source, the mean difference between the dry deposition estimates, using NH3 concentrations measured independently at 0.1m and 0.5m above the canopy, was 2%.Use of a constant surface resistance, with no concentration dependence, as commonly used in inferential models of dry deposition, would have predicted deposition up to eight times too large.