Triple-negative breast cancer (TNBC) is one of the breast cancer subtype that displays a high risk of early recurrence and short overall survival. Improvement of the prognosis of patients with TNBC requires identifying a predictive factor of recurrence, which would make it possible to provide beneficial personalized treatment. However, no clinically reliable predictive factor is currently known. In this study, we investigated the predictive factor of recurrence in TNBC using matrix-assisted laser desorption/ionization-imaging mass spectrometry for lipid profiling of breast cancer specimens obtained from three and six patients with recurrent and non-recurrent TNBC, respectively. The signal for phosphatidylcholine (PC) (32:1) at m/z 732.5 was significantly higher in the recurrence group compared to the non-recurrence group (P = 0.024). PC (32:1) was more abundant in the cancer epithelial area than it was in the surrounding stroma, suggesting that abnormal lipid metabolism was associated with malignant transformation. Our results indicate PC (32:1) as a candidate predictive factor of TNBC recurrence. A future prospective study investigating whether personalized therapy based on PC (32:1) intensity improves the prognosis of patients with TNBC is recommended.