Trade is essential for countries development. In Brazil, coffee has been one of the most important export items and by large is commercialized as a green or roasted bean. The aim of this article is to establish a prediction mode for coffee exports using time series and trend lines. To do so, we collected the exportation volume from the two main export countries in each segment: Brazil and Germany. A five-year forecasting was produced using regression curves provided by Microsoft Excel. Our results indicated that polynomial fits best and this function is consistent with agricultural production that is conditioned to edaphoclimatic factors.