Background To date, there is no approved blood-based biomarker for breast cancer detection. Herein, we aimed to assess semaphorin 4C (SEMA4C), a pivotal protein involved in breast cancer progression, as a serum diagnostic biomarker. Methods We included 6,213 consecutive inpatients from Tongji Hospital, Qilu Hospital, and Hubei Cancer Hospital. Training cohort and two validation cohorts were introduced for diagnostic exploration and validation. A pan-cancer cohort was used to independently explore the diagnostic potential of SEMA4C among solid tumors. Breast cancer patients who underwent mass excision prior to modified radical mastectomy were also analyzed. We hypothesized that increased pre-treatment serum SEMA4C levels, measured using optimized in-house enzyme-linked immunosorbent assay kits, could detect breast cancer. The endpoints were diagnostic performance, including area under the receiver operating characteristic curve (AUC), sensitivity, and specificity. Post-surgery pathological diagnosis was the reference standard and breast cancer staging followed the TNM classification. There was no restriction on disease stage for eligibilities. Results We included 2667 inpatients with breast lesions, 2378 patients with other solid tumors, and 1168 healthy participants. Specifically, 118 patients with breast cancer were diagnosed with stage 0 (5.71%), 620 with stage I (30.00%), 966 with stage II (46.73%), 217 with stage III (10.50%), and 8 with stage IV (0.39%). Patients with breast cancer had significantly higher serum SEMA4C levels than benign breast tumor patients and normal controls (P < 0.001). Elevated serum SEMA4C levels had AUC of 0.920 (95% confidence interval [CI]: 0.900-0.941) and 0.932 (95%CI: 0.911-0.953) for breast cancer detection in the two validation cohorts. The AUCs for detecting early-stage breast cancer (n = 366) and ductal carcinoma in situ (n = 85) were 0.931 (95%CI: 0.916-0.946) and 0.879 (95%CI: 0.832-0.925), respectively. Serum SEMA4C levels significantly decreased after surgery, and the reduction was more striking after modified radical mastectomy, compared with mass excision (P < 0.001). The positive rate of enhanced serum SEMA4C levels was 84.77% for breast cancer and below 20.75% for the other 14 solid tumors. Conclusions Serum SEMA4C demonstrated promising potential as a candidate biomarker for breast cancer diagnosis. However, validation in prospective settings and by other study groups is warranted.