Nitrogen is the most limiting nutrient for smallholder maize production in sub-Saharan Africa (SSA), yet on-farm nitrogen use efficiency remains far below agronomic potential despite decades of investment in fertiliser access. Part of the reason, this paper argues, is that recommendations across the region are built almost entirely around how much nitrogen to apply and say very little about when to apply it relative to crop growth stage and rainfall. That silence on timing is not accidental. It runs through the way agronomic knowledge is produced, packaged, and delivered: trials hold timing constant while varying rate, extension services favour a single scalable rate message, and subsidy programs treat nitrogen as a quantity to be distributed and counted. Under the rain-fed conditions that cover more than 95% of SSA cropland, intra-seasonal rainfall aligned with growth stages can explain 78 to 83% of yield variability, at times outweighing application rate. The result is chronically low nitrogen use efficiency, unreliable fertiliser profitability, and a feedback loop that keeps yield gaps open. None of this requires new technology to fix. Trials need to test timing alongside rate, extension messages need to say when as well as how much, and recommendation tables need a timing column anchored to crop growth stage rather than a footnote.