Abstract A total of 109 cows calving in July and August in a breed × stocking rate trial at Ruakura were blood‐sampled for plasma magnesium (Mg) on seven occasions from June 1990 to March 1991. The objectives were to monitor the effects of month, breed, stocking rate, age, calving day, and covariation with yield of milksolids on Mg concentration. The Friesian and Jersey cows were sired by high breeding index New Zealand Dairy Board bulls, and low and high stocking rates were 3.02 and 3.98 cows/ha respectively for Friesian and 3.57 and 4.53 cows/ha for Jersey. Mg supplementation was applied before calving using MgO on pasture, hay, or silage. Once‐daily drenching with 10 g Mg/cow (as MgCl2) was carried out after calving and continued until November, before the fourth monthly sample was collected. Monthly mean Mg concentrations were 0.87, 0.91, 1.12, 1.06, 0.98, 0.94, and 0.93 mM/1 in June, August, October, December, January, February, and March. There were significant effects of breed in August, October, and overall, with higher Mg concentrations recorded in Jerseys than Friesians. Over all samplings, stocking rate was significant (P < 0.05), with a 4% lower concentration at the high than low stocking rate. In the two months when age was significant (December and January), cows of 5 years and older had lower Mg concentrations than younger cows. Date of calving was not a significant effect. The between‐cow repeatability of Mg concentration from month to month was 0.31 ± 0.04. The regressions of Mg concentration on yields of milk or milksolids were significant for February and March samples, but the regressions only accounted for a small percentage of Mg variation. In conclusion, for four herds supplemented with Mg from pre‐calving until November, there were only small differences in Mg concentration across management groups or breeds, and associations with yield were of minor importance.