Objective To explore the value of combined analysis of speech and electrocardiographic (ECG) features in the early screening of mild cognitive impairment. Methods A convenience sampling method was used to recruit 365 elderly participants from communities in Qixia District, Nanjing, between July and September 2024. Participants were classified into a mild cognitive impairment group and a cognitively normal group based on diagnostic criteria for MCI. Speech data were collected using the “cookie theft picture” task, while ECG signals were recorded simultaneously. The collected data were preprocessed, and features were extracted. Multivariate logistic regression analysis was performed to investigate the relationship between speech and ECG features and the occurrence of MCI in elderly individuals. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to assess the predictive value of speech and ECG features for MCI in the elderly. Results There were 253 in the mild cognitive impairment group and 112 in the cognitively normal group. The results of multivariate logistic regression analysis indicated that features such as long pauses/total speech duration, loudness_sma3_percentile50.0, match score, HRV_IQRNN, and HRV_HTI were significantly associated with the risk of MCI. The ROC analysis demonstrated that the combination of speech and ECG features showed good predictive performance for MCI, with an AUC of 0. 728.Conclusion Combined analysis of speech and ECG features provides a promising approach for the early detection of mild cognitive impairment, offering potential for more efficient and non-invasive screening in clinical practice.