In recent years the growth of air cargo has accelerated, necessitating the assessment of the challenges that this growth will in the future, in order to overcome them and continue contributing to the economic development of the country.Considering the above, this paper proposes several models to estimate demand for air cargo in Colombia, obtained through the use of methodologies such as linear regression and neural networks, which can be used to characterize the current demand and to forecast future demand scenarios given certain contexts to be set.For the estimation of the models presented, information airfreight demand of the top 19 airports in the country was used (in terms of cargo shipped), registered by the Special Administrative Unit of Civil Aeronautics in Colombia (Aerocivil) from 2005 to 2014, besides socioeconomic information on the areas of influence of such airports, obtained mostly by the Administrative Department of National Statistics (DANE) of the same country.Finally a comparison between the results obtained by each modeling methodology, finding better results with neural network models is established.