With the high cost of nitrogen (N) fertiliser – about 20 per cent of variable costs – cotton growers seek to optimise its efficiency. This project aims to improve soil and crop sensor use in N management, and to find a way to automate collecting and analysing data, formulating options and implementing decisions, in order to improve on-farm N use efficiency. By focusing on site-specific management models, it aims for better use of various data: soil, crop and climate; historic on-farm; external public and private, as well as automating decision rules in software that could be linked to real-time application equipment. Several different proximal and remote sensor technologies have been trialled to develop models of N-responsiveness in cotton crops under a range of management conditions.