The main objective of this paper is to present a yield monitor specifically designed for the implementation of precision agriculture in sugar cane crops. The overall system includes a mass flow sensor, a global positioning system receiver, and a data acquisition system. The concept for the sensor is based on the sugar cane yield monitor sensor developed previously at The State University of Campinas. A modified version of the sensor was designed to address the specific needs of a harvesting system, including several sensors to eliminate signal interference, and to monitor the harvesting operation. Field testing of the sugar cane yield monitor system was done during the 2004–2005 harvesting season. The system was mounted on a Case 7700 sugar cane harvester. In order to evaluate the accuracy of the system, each loaded truck was weighed on an electronic platform scale upon arrival at the sugarcane mill. The results showed that the output of the sugar cane yield monitor and the harvested load weight present a correlation of 0·66, and the system performance was stable and reliable during tests. A sugar cane yield map was generated in a 43 ha area plot of a sugar plantation, located in Araras, the state of Sao Paulo, Brazil. The results showed that the yield monitor designed and evaluated is accurate, with a mean error of 4·3% where the maximum error never exceeds 6·4%, and is capable of measuring the yield variability.