A 120-day field experiment was conducted to investigate the responses of soil N2O emissions, plant biomass N content, and N cycling-related functional genes to yak (Bos grunniens) dung pat size, including full-size dung pat (FDP), 1/4FDP, 1/8FDP, and 1/16FDP (i.e., FDP split into four, eight, and sixteen equal-sized dung pat fragments) in an alpine steppe on the Qinghai-Tibetan Plateau. The yield-scaled and cumulative N2O emissions were lower in the 1/16FDP and 1/8FDP than in the FDP and 1/4FDP treatments. In addition, the 1/16FDP treatment had the smallest N2O emission factor (0.002%), possibly due to lower denitrification as shown by the lower nisS, nirK, and nosZ gene copy numbers in the first (day 30) and second (day 72) samplings, and increased aboveground plant N concentration and content, which was 23–32% and 21–36%, respectively, greater than in the other treatments. In conclusion, splitting the yak dung pat into 1/16 fragments would be an effective strategy for managing yak dung to reduce N2O emissions and improve aboveground plant biomass N content which enhances the sustainability of alpine steppe ecosystems on the Qinghai-Tibetan Plateau. The implication from this study is that long-term field experiments should be conducted to further investigate the potential antagonistic or synergistic effects of yak dung pat fragmentation combined with other amendments (e.g., nitrogen inhibitors, lime, or biochar) on reducing N2O emissions.