Winter legume cover crops are a key component of nutrient management in organic agroecosystems. Accurate estimates of N fixation are important for determining the balance of N inputs and outputs and the potential for N loss from these systems. Our first estimates of N fixation for Vicia faba (bell beans) and Vicia dasycarpa (woollypod vetch) using the ¹⁵N natural abundance method were unexpectedly high across all field histories and fertility levels, e.g., 70-98 %Ndfa for vetch, using oats (Avena sativa) as the reference crop. We also observed a negative relationship between d¹⁵N of the non-fixing reference crops and time since last compost addition but no relationship between legume d¹⁵N and compost history, indicating different N pool use by the legumes and reference crop. In a greenhouse experiment, we measured active nodule weight as a proxy for fixation activity to ask if fertility had an effect on early crop N fixation. Preliminary analysis showed active nodules in early bell bean and vetch growth (3.5 wk for bell bean, 2.5 wk for vetch), independent of fertility levels (0, 75 and 200 kg N/ha added to field soil). Evidence of early fixation activity may support the validity of the high %Ndfa estimates. We also present data from a second greenhouse experiment, in which we determined %Ndfa by ¹⁵N natural abundance for vetch grown in field soil with six different management histories and four fertility treatments. These results may help us further determine the potential for high N fixation even under high available soil N conditions.