Background Subclinical acute rejection (subAR) can only be diagnosed by protocol biopsy and is correlated with worse graft outcomes. However, noninvasive biomarkers of subAR are lacked for kidney transplantation recipients in clinic. This study aims to utilize to construct a peripheral blood-based gene signature for subAR diagnosis after kidney transplantation. Methods After systematically screening databases, two cohorts of high quality with 3-month blood profiles and biopsy-proven graft status from the Gene Expression Omnibus databases were employed as training and validation cohorts. Then, the support vector machine recursive feature elimination (SVM-RFE) and the least absolute shrinkage and selection operator (LASSO) logistic regression were used to identify key biomarkers for subAR. Subsequently, the stepwise logistic regression method was applied to construct a gene signature for subAR in the training cohort. Patients were divided into high-risk and low-risk groups based on the cutoff point identified by the receiver operating characteristic (ROC) curve. Then, the signature was validated in a validation cohort with fixed formula. The single-sample gene set enrichment analysis was used to estimate immune cells in the blood. Results Fifty key biomarkers were filtered out with the machine learning algorithms. Then, a novel six-gene signature was constructed using the LASSO and stepwise logistic regression method. The signature had high accuracy in both training [area under the curve (AUC) =0.923] and validation cohort (AUC =0.855). Additionally, these six genes were found to have significant and consistent relationships with blood immune cells in both cohorts, especially for T cells subtypes. Conclusions We developed and validated a novel noninvasive six-gene signature based on peripheral blood to diagnose subAR, which offered a potential tool for clinical practice. The six-gene signature offered a potential method to monitor patients following transplantation and make a timely intervention.