N,N-dimethylsphingosine (DMS) has been documented to be in vitro protective against myocardial ischemia–reperfusion injury (IRI) and can recruit CD4⁺CD25⁺Foxp3⁺ regulatory T cells (Tregs), which may participate in the cardioprotection. We hypothesized that when in vivo applied after a myocardial ischemia, DMS may be cardioprotective by recruiting Tregs. Myocardial IRI was induced in C57BL/6 mice by occluding the left main coronary arteries followed by relaxation, and DMS (0.43 mg/kg) was intravenously injected 5 min after the onset of ischemia. We found that in wild-type (WT) mice, compared with the ischemia–reperfusion group, DMS reduced the infarct size (47.1 ± 8.9 vs. 33.1 ± 3.4 %, p < 0.01), and neutrophil infiltration at 24 h reperfusion (R) evaluated by TTC and immunohistochemical staining, respectively, and increased the aggregation of Tregs [(6 ± 1)/mm² vs. (30 ± 4)/mm², p < 0.01], peaking at 1 h R by immunofluorescence staining, with reduced gene expression of inflammatory factors at 4 h R in the reperfused myocardium by real-time PCR. This protection was abolished by phosphatidylinositol 3-kinase (PI3K)/Akt inhibitor or Tregs-depleting antibody. Relative to WT mice, the cardioprotection conferred by T cell- and B cell- deficient Rag2 knockout (KO) mice was not strengthened by DMS or by DMS and the adoptive transfer of Tregs from WT mice, but was abolished by DMS and WT non-Tregs and was recaptured by the cotransfer with WT Tregs but not with Akt1^+/− mice-derived Tregs. In conclusion, applied at an early stage of ischemia, DMS may be in vivo protective against myocardial IRI by recruiting Tregs via PI3K/Akt pathway.