Brazil is world leader in the production of sugarcane, crop that is grown in several regions of the country and responsible for the generation of sugar and ethanol. However, maintaining high productivity in the field requires that productive areas are increasingly equipped with innovative solutions that provide increased efficiency in crop management. The main focus of agricultural systematization is using georeferencing and geoprocessing to make more efficient the trafficability of agricultural machines and increase the productive area, reducing the costs with field operations and the risks of failures in cultures caused by trampling. The objective of this study was to evaluate the operational efficiency of autopilot on trafficability and reduction of trampling into a sugarcane harvesters cultivated area. Efficiency tests were carried out in an area cultivated with sugarcane divided between two treatments, using the autopilot with satellite navigation coupled in the harvester and in the other area, the harvester maintained the common working method (without use of automatic pilot). The data generated by the equipment were collected and the parameters of fuel consumption, operational time, production harvested and harvest losses were evaluated, compared and interpreted through multivariate statistical analysis. The results of this study indicated that the treatment using harvester with autopilot, there was significant field harvest efficiency (7.6 Mg h-1) and a reduction in fuel consumption (on average 10%) when compared to the harvester without autopilot.