Precision horticulture and site–specific weed management (SSWM) practices based on remote sensing techniques have a great potential to contribute to the development of intelligent, effective, and more environmentally friendly weed management systems. The most crucial and challenging task to accomplish is correct weed identification in the field. Non–imaging and imaging ground-based sensors can be used for this purpose. Non–imaging weed identification systems include spectrometric sensors, optoelectronic sensors, UV–induced fluorescence sensors, LiDAR sensors, and ultrasonic sensors. As for imaging systems, RGB–imaging is the most widespread remote sensing technique to detect weed patches in the field while the combination of imaging and non–imaging sensors to obtain hyperspectral data is also very promising. Unmanned aerial vehicles (UAV) equipped with smart sensors can also be used to collect information on weed flora composition in horticultural systems. Once data have been captured, they are processed with machine learning algorithms to distinguish between either weeds and bare soil or between weeds and crops. Data can be used to create weed maps to guide precise mechanical weed control operations and site-specific herbicide applications. Sensor–based mechanical weed control can be achieved with semi-autonomous tractor–drawn implements and autonomous robots. To perform site-specific herbicide applications, a reliable weed identification system, a processing unit, and precision weed control implements are needed. Precision horticultural methodologies can result in significant herbicide savings while sensor-based methods can be used to evaluate herbicide efficacy. However, the use of site-specific weed management tools by growers remains sporadic. To increase adoption rates, technical experts need to establish stronger links with growers and their advisers to familiarize them with the concepts of precision horticulture and site-specific weed management.