The goal of this paper is to advance in the current understanding about the impact of innovation (in this case defined as investment in innovation activities) on the non-traditional exports. The study analyzes a dataset of Colombian companies that perform their activities in the sectors of the International Standard Industrial Classification - ISIC - from 2005 to 2012. We use a panel data model in which through Box Jenkins theory, is able to identify statistically significant variables on export performance. The findings allow verify theories about the positive relationship between these variables, and in our particular case to demonstrate the impact of innovation activities on export development. Finally, the results suggest that the stimulation of innovation and policies related to this are essential for the growth of exports.