This working paper explores a generic method that can be used to benchmark nitrogen (N) input requirements for crop production and the efficiency by which inputs are used. Two types of N benchmarks are introduced: one for short-term and another for long-term assessments. We explain the underlying assumptions, data requirements and types of applications. Both benchmarking methods are especially suitable for regional, national or global analyses. The proposed methodology is illustrated for cereal production (maize, wheat, rice, millet and sorghum) in ten countries in sub-Saharan Africa, under current and optimal nutrient management, for today and towards 2050. We show that agronomic nitrogen-use efficiency (NUE) can be two to four times larger than currently observed in on-farm trials for the long- term benchmark. Potential improvements in N input requirements are related to greenhouse gas (GHG) emission mitigation potentials, using scenarios that include population increase and dietary change, potential yield increase and avoided land reclamation. Here, we show that when following the current trajectory of yield trends while maintaining the low current nitrogen-use efficiency, GHG emissions from cereal production will be three times larger than sustainable intensification of cereals in sub-Saharan Africa. The proposed N benchmarking method is most useful for regional or larger scale analyses and less useful for field assessments. Nonetheless, this might fill a gap in higher scale analyses, especially for estimating potential improvements in NUE and reducing GHG emissions. This working paper presents work in progress. In the future, we will test the proposed methodology on different case studies to evaluate its potential and finetune its operation.