Understanding the influence of management practices on greenhouse gas (GHG) emissions from greenhouse cropping is of great importance for assessing the environmental impacts of the greenhouse cropping industry and improving its sustainability. A tomato–cucumber–tomato rotation experiment was carried out in a typical solar greenhouse in northwest China with four treatments including two irrigation levels (regular (RI) and low (LI)) and two fertilizer types (organic fertilizer (N1) and inorganic fertilizer (N2)). Greenhouse gas fluxes (nitrous oxide, N₂O, and methane, CH₄) were measured regularly using the closed chamber method during the three growing seasons, along with soil water-filled pore space (WFPS), temperature, mineral-N concentration and nitrobacteria, nitrosomonas and denitrifying bacteria abundance. Over the rotation, the soil acted as a source for N₂O and a sink for CH₄, with the mean fluxes of 0.12 mg N₂O-N m⁻² h⁻¹ and -0.31 mg CH₄-C m⁻² h⁻¹, respectively. The stepwise multiple linear regressions indicated that WFPS and soil temperature accounted for significant portion of N₂O emission and CH₄ uptake variations, respectively for both fertilizer types. Fertilization rate and type resulted in much greater difference of cumulative GHG emission between treatments than the irrigation level. Inorganic fertilizer with higher nitrogen application rate usually resulted in higher cumulative N₂O emission and lower CH₄ uptake than organic fertilizer application. Over the rotation, total greenhouse emission (GHGt) and greenhouse emission intensity (GHGI) on average followed the same order of RIN2 > LIN2 > LIN1 > RIN1 with N₂O emission as the dominant component for each treatment. Overall, organic fertilizer with proper water application under drip irrigation can effectively mitigate greenhouse gas emissions and maintain relatively high and stable vegetable yields in solar greenhouse cropping in northwest China.