Alzheimer’s disease (AD) is a neurodegenerative disorder that affects millions of people around the world with no available treatments to cure, reverse or even slow down the disease progression. Detection of AD in the early stages, before the massive neuronal loss, is key to developing disease-modifying treatments and preventive strategies. The current AD diagnosis is largely based on evaluation tests such as Mini-Mental State Examination (MMSE), imaging techniques such as Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET), and postmortem autopsies. It is expensive and only available at the later/severe stages when the brain of the patient is already damaged or after the death of the patient. Cerebrospinal fluid (CSF) tests have the ability to identify earlier stages with the help of certain biomarkers but large scale use is often limited by invasive CSF collection protocols. This study is a part of Blood Biomarker-based diagnostic tools for Alzheimer’s disease (BBDiag) network. It is aimed at developing graphene based electrochemical biosensors for the detection of blood biomarkers as blood sampling is far less complex, minimally invasive, and cheaper compared to CSF sampling. Electrochemical biosensors are extremely advantageous for biomarker detection. This is due to their rapid response (fast analysis with results available in a few minutes), portability (home testing of diseases), cost-effectiveness (inexpensive analysis which allows the possibility of large scale implementation, particularly to low-income communities with no access to sophisticated instrumentation such as MRI/PET scanners), easy handling (user friendly with easy to understand results) and disposability (single use sensors with no requirement for maintenance). Using graphene as a base material helps in improving the sensitivity of these biosensors as the absence or presence of very few analyte molecules can trigger a recognisable change in its electrical properties. In this work, two label-free graphene based electrochemical biosensors have been developed and validated with well-known biomarkers, Aβ. The first biosensor is a graphene/ reduced graphene oxide (rGO) dual-layer biosensor for the detection of Aβ1−42. The dual-layer is obtained by modifying graphene screen printed electrodes (SPEs) with electrochemically reduced rGO. The dual-layer is further modified with 1-pyrenebutyric acid N-hydroxysuccinimide ester (Pyr-NHS) linker, which attaches non-covalently via π − π stacking on the surface of rGO and facilitate the immobilisation of antibodies. The surface characterisation is achieved using various techniques including Raman spectroscopy, X-ray photo electron spectroscopy (XPS), scanning electron microscopy and cyclic voltammetry. Differential pulse voltammetry is used to evaluate the analytical performance of the biosensor. The limit of detection (LOD) is found to be 2.4 pM over a linear range of 11 pM to 55 nM. The biosensor depicts high selectivity for Aβ1−42 in the presence of Aβ1−40 and ApoE ε4 species. This is an important requirement for reliable detection from biofluids as these interfering species can be present in excess. The graphene/rGO dual-layer biosensor shows this improvement over existing Aβ biosensors that fail to distinguish between Aβ1−40 and Aβ1−42. It was employed for the detection of spiked human and mice plasma samples. The sensing results obtained from an age-based study of mice samples revealed a decrease in the plasma levels of Aβ1−42 with a progression of AD from 9 months to 12 months. This is correlated to the increased Aβ plaques in the brain of 12 months old mice as revealed by immunohistochemistry and magnetic resonance imaging data. The second biosensor is based on an amine (NH2) functionalised rGO SPE for the detection of both Aβ1−40 and Aβ1−42. NH2 linkers are predominantly attached on the edge and defect sites of rGO SPE via chemisorption as revealed by XPS, FTIR, and Raman analysis. LOD of the biosensor is calculated to be of 9.51 fM over a linear range of 10 fM-10 pM for Aβ1−40 and 8.65 fM over a linear range of 10 fM-50 pM for Aβ1−42. This is the lowest reported LOD by a label-free graphene biosensor. This improvement in sensitivity is attributed to higher antibody binding sites on the surface provided by the NH2 linker. In addition, the biosensor depicts excellent selectivity in the presence of interfering Aβ and ApoE ε4 species. It is also successfully validated with spiked human plasma within its linear range. Therefore, both graphene/rGO and rGO/NH2 biosensors show potential to be developed into a point of care technologies to provide rapid, sensitive, and selective detection of blood-based AD biomarkers. Due to the difference in the sensitivity of the two biosensors, they can be applied for the detection of different biomarkers (depending on their concentration in plasma) or one biomarker in different stages of disease progression. However, further validation with clinical samples is needed before they can be developed into commercial devices for minimally invasive and time-efficient routine screening of AD.