Flax fertility trials were conducted over a three-year period at eight locations, primarily in Saskatchewan but also in Alberta and Manitoba. All fertilizer was side-banded, and the treatments were a factorial combination of four nitrogen (N) fertilizer rates (13, 50, 100, and 150 kg N/ha) and four phosphorus (P) fertilizer rates (0, 20, 40, and 60 kg P₂O₅/ha) arranged in an RCBD with four replicates. The response variables measured were plant density, days to maturity, seed yield, and test weight. Residual NO3-N was rather variable across locations; however, P levels were relatively low with a maximum of 24 ppm (Olsen P) and less than 10 ppm at 63% of the sites. Flax emergence was somewhat sensitive to side-banded urea whereby stand reductions associated with increasing N rate were observed at 74% of the sites. Amongst the affected sites, the response was linear with a 28% reduction in plant densities when the N rate was increased from 13 kg N/ha to 150 kg N/ha. Side-banded monoammonium phosphate did not affect plant density, regardless of rate. Increasing N rate delayed maturity 71% of the time; however, averaging 2.4 days amongst the affected sites this is not expected be of practical concern under normal conditions.  Phosphorus rate did not have a noticeable effect on flax maturity. Flax yields were increased with both N and P fertilizer. Focussing on N, there was a site by N rate interaction whereby the response was relatively strong at 83% of the sites, increasing yields by 39% on average with maximum yields achieved at approximately 100 kg N/ha. At the remaining sites, the response was weak with an 11% yield increase on average and optimal rates closer to 50 kg N/ha. For phosphorus, although there was variation, no site by phosphorus interaction was detected. The average response was linear but relatively shallow (7%); therefore, more modest rates of 20-40 kg P₂O₅/ha are likely to be most economical and enough to maintain soil fertility under most circumstances. At 50% of the sites, the maximum yield increase with P was 5-10% while the response was below 5% at 28% of the sites and greater than 10% at 22% of the sites. Test weight was not affected by P fertilizer rate but there was a very slight linear increase at 41% of the sites. In conclusion, these results show that adequate N and P fertility are both important for achieving higher flax yields; however, the responses were modest with respect to both magnitude of the yield increase and the rates at which maximum yield was achieved. Although this in not necessarily unexpected with any crops, site-to-site variability was much higher than the variability within sites due to N and P fertilizer rate. This potentially suggests that fertility is not likely the most limiting factor for majority of western Canadian flax acres; however, this will vary on a farm-to-farm basis.

• Flax emergence was sensitive to side-banded urea whereby there were linear reductions in plant density with increasing N rate at approximately three-quarters of the site-years and, amongst these, an average plant loss of 28% as the rate was increased from 13 kg N/ha to 150 kg N/ha. Side-banded monoammonium phosphate did not affect flax emergence.
• Increasing N rate led to slight but significant delays in maturity approximately 70% of time. The average delay amongst the affected site-years was 2.4 days when the N rate was increased from 13 kg N/ha to 150 kg N/ha – this would not be expected to be of any agronomic concern under most circumstances. Phosphorus fertilization did not have any noteworthy effects on maturity.
• Flax yield increased with N fertilization in essentially all cases; however, the response was weak at 17% of the site-years where either residual N was high, or yields were more limited by other environmental factors. Amongst the more responsive sites, yields plateaued at approximately 100 kg N/ha and the magnitude of the yield increase averaged 39%. At the less responsive sites the average magnitude of response was 11% and there was little benefit to fertilizer rates exceeding 50 kg N/ha.
• At 7% when averaged across site-years, the magnitude of the flax yield response to P fertilizer was smaller than that observed for N; however, it did appear to be reasonably consistent. The response was linear, but the slope was likely too shallow to justify rates exceeding 20-40 kg P₂O₅/ha. Amongst individual sites, the magnitude of the individual responses to P ranged from nil to nearly 40% with responses between 5-10% at half the sites, below 5% at 28% of the sites, and greater than 10% at 22% of the sites.
• Flax test weight was affected by N rate at 41% of the site-years where there was a slight linear increase in test weight with N rate. The effect was small with, specifically amongst the responsive sites, only a 2% increase in test weight at the 150 kg N/ha rate relative to the 13 kg N/ha rate.