Abstract: As environmental concerns grow globally, many countries are elaborating upon a plan to reduce greenhouse gas emissions which can result in global climate change. Cattle production is one of the recognized sectors in agriculture that produce a large amount of methane from enteric fermentation, one of the major greenhouse gases being targeted for reduction. Enteric methane production by cattle varies between 2-12% of gross energy intake and a recent statistics showed that it contributes >20% of the total methane emissions in the US dairy cattle is known to produce more enteric methane than beef cattle due to a relatively large amount of forage in the diet and a high level of intake. Therefore, reducing methane emissions by dairy cattle has become one of the most important areas of research in the modern agriculture and accurate quantification of methane emissions by dairy cattle is critical. Direct measurement of methane emissions by dairy herds requires a large amount of time, labor and money and it cannot be practically used to estimate methane emissions from each farm. Application of modeling to predict methane emissions thus could be an alternative and better way of quantifying methane emissions from dairy herds. A common modeling approach is to develop a methane emission model empirically which is heavily dependent on statistical analysis on available data. An Empirical Model is very useful and its predictability may be satisfactory as long as it is built from sufficient and appropriate accumulated data. Interpolation beyond the range of data should be avoided. Many published models can be classified as Empirical Models. A Mechanistic Model, on the contrary, emphasizes more on the underlying mechanism. Experimental data are only used for parameterization of the variables and evaluation of the model. In many cases a Mechanistic Model requires more variabes to be estimated than an Empirical Model which may limit its versatile use. One important feature of a Mechanistic Model is that unlike an Empirical Model it can be easily modified and applied to different conditions (climate, feedstuff, breed and management) without changing the structure of the model. A relatively small number of Mechanistic Models have been published. Each type of models has its pros and cons and one should thus be cautious when choosing a model for a specific condition. According to the model comparisons in literature, the overall predictability of the published models is still low and needs to be improved with further research. More accurate predictions of methane emission by dairy cattle require the development of a more mechanistic model that accounts for more of the biologically important variables that affects methane emissions and this model should be able to integrate all of the farm-specific components. It can be concluded that modeling is very useful to predict the methane emissions by dairy cattle and it is also helpful to find the most appropriate mitigation strategy for a specific condition.