The expansion of solar photovoltaic (PV) infrastructure into semi-arid grasslands has the potential to alter carbon (C) cycling processes. We conducted a three-year field experiment in Colorado, USA to investigate the effects of PV panel microclimates and ecovoltaic management practices (interseeding native species and irrigation) on above and belowground net primary productivity (ANPP and BNPP), soil organic carbon (SOC) stocks, and soil CO2 flux. Total SOC stocks from 0-30 cm did not differ between baseline and 3 years after PV installation, but the microclimate directly under the PV panels (Under) had greater SOC and standing root biomass, and lower soil respiration rates than areas between rows (Open) or on the panel edges of the PV systems. Interannual precipitation variability modulated PV microclimate effects, with the Open microclimate having significantly higher BNPP during the wettest year and the Under microclimate having significantly lower BNPP during the driest year. The Under microclimate also promoted a deeper root distribution, potentially enhancing subsoil C inputs. Irrigation had limited effects on ANPP and BNPP, but increased soil CO2 flux. Interseeding native plants did not significantly influence C cycling compared to maintaining existing pasture vegetation. Our findings highlight the complex interactions between PV infrastructure, microclimate, and precipitation in driving grassland C cycling in the years immediately following installation. Low-impact PV installation methods that preserve vegetation and minimize disturbance appear crucial for maintaining soil C sequestration services. Ecologically-informed PV array design and management strategies may help optimize C cycling co-benefits in semi-arid grasslands and longer-term studies across diverse sites are needed to fully understand the environmental impacts of this emerging land use.