The complex triplet potential energy surface of the CH2N2O system, including 49 minimum isomers and 114 transition states, is investigated at the B3LYP and QCISD(T) (single-point) levels in order to explore the possible reaction mechanism of the (CH2)-C-3 radical with N2O. The most feasible pathway is the head-on attack of (CH2)-C-3 at the terminal N-atom of N2O to form cis-H2CNNO (a(1)) and trans-H2CNNO (a(2)). Both a(1) and a(2) can subsequently dissociate to give P-1 (H2CN + NO) via the direct N-N bond rupture. Much less competitively, a(1) can undergo a 1,4-H shift, leading to the chainlike isomer HCNNOH (k(1)), followed by the direct N-N bond cleavage to form product P-2 (HCN + (HON)-H-3) or interconversion between the isomers k(1)-k(8) and subsequent dissociation to P-2. Furthermore, the products P-2 (H2CN + NO) and P-2 (HCN + (HON)-H-3) can undergo secondary dissociation to the same product P-12 (HCN + NO + H). The formation of CO, however, seems impossible due to rather large barriers. Our results are in part contradictory with the recent time-resolved Fourier transform infrared spectroscopic study that nascent vibrationally excited products CO, NO, and HCN were observed. Since the initial N-attack step from R to a(1) needs a considerable barrier of 14.8 kcal/mol, the title reaction may only be significant at high temperatures, as confirmed by the ab initio dynamic calculations on the rate constants. The reactivity discrepancies between the triplet and singlet CH2 with N2O are compared and discussed in terms of their potential energy surface features. Our calculations suggest that future experimental reinvestigations on the product distributions and rate constants of the title reaction at high temperatures are greatly desired.