A reliable quantification of nitrogen (N) losses by ammonia (NH 3) volatilization can contribute to identifying optimized strategies of fertilizer management. The objective of this study was to quantify ammonia volatilization from several organic N sources incorporated into or applied onto the soil surface. Two cultivation areas, under snap bean and corn, were evaluated at Embrapa Agrobiology, Seropédica, Rio de Janeiro (RJ). Both experiments used a randomized complete block design in split-plots, with four replications. The main plots consisted of four organic fertilizers (castor bean cake, bokashi, legume fertilizers, cattle manure), at rates of 200 kg ha-1 N, and a control treatment (without fertilization), and the subplots corresponded to the management forms (incorporated or surface-applied) of the fertilizers. In the first experiment, snap bean cv. Novirex was grown in winter/spring and in the second, corn cv. Catingueiro in summer/autumn. In each subplot, static semi-open NH 3 collectors were installed. We conclude that surface-applied castor cake was the organic fertilizer with highest N loss by NH 3 volatilization. A comparison of the management systems (incorporated or surface-applied) showed that volatilization from organic fertilizers incorporated into the soil was significantly lower, with a reduction of 80 % for castor cake, of 78 % for bokashi and 67 % for legume fertilizer. Nitrogen loss through ammonia volatilization varied, from rates of 3 to 25 % in winter/spring and from 2 to 38 % in summer/autumn, according to the organic fertilizer applied. The period required to recover 95 % of the N lost by NH 3-N volatilization was between 13 to 18 days for castor cake; 14 to 43 days for bokashi; 17 to 49 days for legume fertilizer and more than 43 days for cattle manure.