Atmospheric deposition of reactive nitrogen (Nr) has enriched oligotrophic lakes with nitrogen (N) in many regions of the world and elicited dramatic changes in diatom community structure. The lakewater concentrations of nitrate that cause these community changes remain unclear, raising interest in the development of diatom-based transfer functions to infer nitrate. We developed a diatom calibration set using surface sediment samples from 46 high-elevation lakes across the Rocky Mountains of the western US, a region spanning an N deposition gradient from very low to moderate levels (<1 to 3.2 kg Nr ha⁻¹ year⁻¹ in wet deposition). Out of the fourteen measured environmental variables for these 46 lakes, ordination analysis identified that nitrate, specific conductance, total phosphorus, and hypolimnetic water temperature were related to diatom distributions. A transfer function was developed for nitrate and applied to a sedimentary diatom profile from Heart Lake in the central Rockies. The model coefficient of determination (bootstrapping validation) of 0.61 suggested potential for diatom-inferred reconstructions of lakewater nitrate concentrations over time, but a comparison of observed versus diatom-inferred nitrate values revealed the poor performance of this model at low nitrate concentrations. Resource physiology experiments revealed that nitrogen requirements of two key taxa were opposite to nitrate optima defined in the transfer function. Our data set reveals two underlying ecological constraints that impede the development of nitrate transfer functions in oligotrophic lakes: (1) even in lakes with nitrate concentrations below quantification (<1 μg L⁻¹), diatom assemblages were already dominated by species indicative of moderate N enrichment; (2) N-limited oligotrophic lakes switch to P limitation after receiving only modest inputs of reactive N, shifting the controls on diatom species changes along the length of the nitrate gradient. These constraints suggest that quantitative inferences of nitrate from diatom assemblages will likely require experimental approaches.