This 3-year field study evaluated the effects of winter cover crop treatments, including no cover crop (control), elbon rye (Secale cereale L.), daikon radish (Raphanus sativus ssp. acanthiformis), Austrian winter field peas (Lathyrus hirsutus), and their mixture, on topsoil (0-10 cm) aggregate size distribution and stability, as well as soil organic carbon (SOC) and total nitrogen (TN) stocks in bulk soil and aggregate fractions under a no-till dryland cotton-corn rotation. Results revealed the <0.25 mm aggregates were the most dominant (63.6%-74.8%), exhibited the highest C and N preservation capacity due to their high aggregate content, and contributed the most to SOC and TN stocks, accounting for 40.8%-65.9% and 36.1%-65.1%, respectively. No significant differences in aggregate size distribution or stability were found among treatments. The treatment with radish, peas, and rye planted sequentially over 3 years exhibited the highest SOC stock (28.2 Mg ha(-1)), while the peas treatment had the highest TN stock (3.1 Mg ha(-1)) in bulk soil. The radish-radish-mixed (peas + radish + rye) sequence increased the proportion of SOC (38.8%) and TN (42.8%) within the 2-1 mm and 1-0.5 mm aggregate fractions. In the <0.25 mm aggregates, the peas treatment had the highest SOC (11.2 Mg ha(-1)), and the treatment with 2 years of radish followed by a mixed species of peas, radish, and rye had the highest TN (0.8 Mg ha(-1)). The main factors influencing bulk SOC stock were the C preservation capacity and SOC stock of the 0.5-0.25 mm aggregates, while bulk TN stock was primarily driven by the N preservation capacity and TN stock of the >2 mm aggregates. The findings suggest that cover crop strategies involving peas and species diversification, such as sequential planting of radish, peas, and rye, can enhance SOC and TN accumulation, particularly in the <0.25 mm aggregates, thereby improving soil health in no-till dryland systems.