Rice-shrimp rotation systems are one of the widespread farming practices in the Vietnamese Mekong Delta coastal areas. However, greenhouse gas (GHG) emissions in the system have remained unclear. This study aimed to examine methane (CH4) and nitrous oxide (N2O) emissions from the system, including (i) land-based versus high-density polyethylene-lined (HDPE) nursery ponds and (ii) conventional versus improved grow-out ponds inoculated with effective microorganisms (EM) bioproducts. The results showed that CH4 flux in land-based and HDPE-lined nursery ponds were 1.04 and 0.25 mgCH4 m−2 h−1, respectively, while the N2O flux was 8.37 and 6.62 μgN2O m−2 h−1, respectively. Global warming potential (GWP) from land-based nursery ponds (18.3 g CO2eq m−2) was approximately 3 folds higher than that of the HDPE-lined nursery pond (6.1 g CO2eq m−2). Similarly, the mean CH4 and N2O fluxes were 15.84 mg CH4 m−2 h−1 and 7.17 μg N2O m−2 h−1 for the conventional ponds, and 10.51 mg CH4 m−2 h−1 and 7.72 μg N2O m−2 h−1 for the improved grow-out ponds. Conventional practices (2388 g CO2eq m−2) had a higher 1.5-fold GWP compared to the improved grow-out pond (1635 g CO2eq m−2). The continuation of the land-based nursery pond and conventional aquacultural farming practices increase CH4 emission and GWP, while applying HDPE-lined nursery ponds combined with improved grow-out ponds could be a promising approach for reducing GHG emissions in rice-shrimp rotation systems. This study recommends further works in the rice-shrimp rotation systems, including (i) an examination of the effects of remaining rice stubbles in the platform on the availability of TOC levels and GHG emissions and (ii) ameliorating dissolved oxygen (DO) concentration on the effectiveness of GHG emission reduction.