Cancer biomarkers may hold the key for the early detection of cancer, distinguishing between benign and malignant cells, and differentiating tumor types. The detection of cancer biomarkers can be used for cancer diagnosis, monitoring the response to therapy, and providing real-time prognostic information for cancer patients. Most of the traditional cancer biomarker detections are based on specific antibodies or expensive instrumentations/complex operations. In this dissertation, we have developed aptamer-based bioassays for visual detection of cancer protein biomarkers with low-cost and short assay time. Aptamers with specific sequences are immobilized on gold nanoparticle (AuNP) surface through self-assembling process. Combining the excellent molecular recognition properties of aptamers and the unique optical properties of AuNPs, colorimetric assay for carcinoembryonic antigen (CEA) and mucin 1 (MUC1) (breast cancer biomarkers) and lateral flow assay for platelet-derived growth factor (PDGF) and thrombin have been developed. The methods were applied to detect protein biomarkers in human plasma and blood successfully. The sensitivities of the assays were further improved by using enzyme-coated AuNP dual labels and designing the cross-flow test strips. The developed approaches have the potential to be extended for detecting other biomarkers, and show great promise for point-of-care or in-field detection.