The success of variable rate N fertilizer application rests on our ability to predict the contribution of soil N to growing crops. We assessed relationships between soil N availability indices (SNAIs), yield, and total N accumulation of wheat ( Triticum aestivum L.) grown in a typical glacial till landscape in Saskatchewan, Canada. Soil samples were collected at 3‐m intervals along a 300‐m transect comprised of low (LCFS) and high catchment footslopes (HCFS), and low (LCSH) and high catchment shoulders (HCSH). Total soil N and C, organic C, mineral N, depth of A horizon, spring soil moisture, grain yield, and total plant N were measured. Soil N availability indices used in this study included: (i) cumulative N released during a 2‐wk aerobic incubation (N MIN ); (ii) potentially mineralizable N estimated using a 16‐wk aerobic incubation (N 0 ); (iii) NO 3 sorbed on anion‐exchange membranes (NO 3AEM ); (iv) N extracted with hot KCl (N KCl ); and (v) N hydrolyzed with hot KCl (N HYDR ). Although all SNAIs were significantly correlated to yield and, with the exception of N 0 , total plant N when analyzed across the transect, typically <40% of the yield variability was explained. Forward stepwise regression revealed that most SNAIs failed to explain more variability in crop N accumulation than did basic soil properties or relative elevation. Although these results do not invalidate the use of SNAIs for soil testing purposes, it is clear that SNAIs must be combined with additional information about field scale variability for predicting fertilizer N requirements. Without this information, grid sampling as a means of assessing N requirements remains ill‐advised for glacial till semi‐arid landscapes.