Environmental indicators for nitrogen (N) use efficiency (NUE) based only on N inputs and N removal are becoming widely used in science, policy, and commercial supply chains to track the sustainability of food production. However, these indicators do not reflect the contribution of inherent soil productivity, which can supply half of crop demand and is therefore a core principle in determining how much fertilizer is needed and the corresponding risk of N losses. Using a global dataset of optimal N rates for crop production, we evaluated the performance of conventional (N recovery efficiency) and simplified NUE indicators to understand their relationship and respective limitations, helping inform policy efforts for simultaneously meeting food production and sustainability goals. A key finding is that conventional agronomic approaches designed to optimize crop productivity and profit related to N fertilizer inputs have tradeoffs for environmental performance, with only 35 and 31 % of observations (n=448) falling within sustainable ranges for NUE. Meanwhile, simplified NUE indicators such as N balance or the ratio of N outputs to inputs were unable to detect sites with inherently low N recovery efficiency and high risk of N losses, highlighting a weakness of neglecting soil N supply in their calculations. Together these results suggest the need for a combined approach that merges insights from locally available agronomic data on N recovery efficiency with global environmental thresholds for NUE. Using our findings as a case study, we propose new steps forward for evaluating NUE in different cropping systems and regions to enhance food security while mitigating N pollution.