Field application of liquid animal manure (slurry) is a significant source of ammonia (NH ₃) emission to the atmosphere. It is well supported by theory and previous studies that air temperature effects NH ₃ flux from field applied slurry. The objectives of this study was to statistically model the response of temperature at the time of application on cumulative NH ₃ emission. Data from 19 experiments measured with the same system of dynamic chambers and online measurements were included. A generalized additive model allowing to represent non-linear functional dependences of the emission on the temperature revealed that a positive response of the cumulative NH ₃ emission on the temperature at the time of application up to a temperature of approximately 14 °C. Above that, the temperature effect is insignificant. Average temperature over the measuring period was not found to carry any additional information on the cumulative NH ₃ emission. The lack of emission response on temperature above a certain point is assumed to be caused by drying out of the slurry and possible crust formation. This effect is hypothesized to create a physical barrier that reduce diffusion of NH ₃ to the soil surface, thereby lowering the emission rate. Furthermore, the effect of the interaction between soil type and application technique and the effect of dry matter content of the slurry was derived from the model, and found to be significant on cumulative NH ₃ emission predictions.