Ammonia (NH3) and nitrous oxide (N-2O) emissions were measured in the field for three months from three different herbage mulches and from bare soil, used as a control. The mulches were grass with a low N-content (1.15% N in DM), grass with a high N-content (2.12% N in DM) and alfalfa with a high N-content (4.33% N in DM). NH3 volatilization was measured using a micrometeorological technique. N-2O emissions were measured using closed chambers. NH3 and N-2O emissions were found to be much higher from the N-rich mulches than from the low-N grass and bare soil, which did not differ significantly. Volatilization losses of NH3 and N-2O occurred mainly during the first month after applying the herbage and were highest from wet material shortly after a rain. The extent of NH3-N losses was difficult to estimate, due to the low frequency of measurements and some problems with the denuder technique, used on the first occasions of measurements. Nevertheless, the results indicate that NH3-N losses from herbage mulch rich in N can be substantial. Estimated losses of NH3-N ranged from the equivalent of 17% of the applied N for alfalfa to 39% for high-N grass. These losses not only represent a reduction in the fertilizer value of the mulch, but also contribute appreciably to atmospheric pollution. The estimated loss of N-2O-N during the measurement period amounted to 1% of the applied N in the N-rich materials, which is equivalent to at least 13 kg N-2O-N ha-1 lost from alfalfa and 6 kg ha-1 lost from high-N grass. These emission values greatly exceed the 0.2 kg N-2O-N ha-1 released from bare soil, and thus contribute to greenhouse gas emissions.