Abstract Background Oesophageal cancer is the 6th leading cause of cancer death worldwide, accounting for 1 in 18 cancer-related mortality in 2020. High mortality rates with oesophagogastric cancer is largely related to late diagnosis. In England and Wales, 35% of oesophagogastric cancers are diagnosed at stage 3 and 42% at stage 4. Early detection of oesophageal cancer improves survival. Currently, the only reliable and effective modality for early diagnosis is oesophagogastroduodenoscopy (OGD). However, OGD is an invasive procedure that carries risks, such as bleeding and visceral perforation. Thus, there is no national screening programme for oesophagogastric cancer, although patients with diagnosed Barrett's oesophagus undergo endoscopic surveillance. Cytological sampling devices such as Cytosponge and EsophaCap are less invasive but still require ingestion of a medical device, and are not without risks, including detachment of device, pharyngeal bleed, and failure to swallow. The discovery and validation of blood biomarkers for the screening and diagnosis of oesophageal cancer may provide advantages of non-invasive, rapid, and cost-effective testing. Yet, currently, there is no clinically validated blood screening biomarker. In this systematic review, we aim to identify developing blood biomarkers that have been described in the literature for early diagnosis of oesophageal cancer. Methods A systematic search was conducted on EMBASE (1974-), Medline (1947-), and Web of Science (1980-) from inception to 3rd November 2021. Exploded subject headings and combinations of keywords were used. Search terms used were [oesophageal cancer or oesophageal ca or oesophageal malignancy or oesophageal neoplasm or Barrett's oesophagus or Barrett's metaplasia] AND [Tumour biomarkers or tumour markers or biological marker] AND [diagnosis or early diagnosis or screening or detection or discrimination] AND [adenocarcinoma or squamous cell carcinoma] AND [blood or serum or plasma] AND [esophagus or oesophagus]. Papers describing blood biomarkers that hold potential in screening and diagnosing oesophageal cancer were included. We focussed on adenocarcinoma, and squamous cell carcinoma affecting the oesophagus and the oesophagogastric junction. Tumour markers with a role in more than one type of cancer were included. We excluded rarer histological types. Biomarkers utilised to prognosticate outcomes of therapy were excluded, as were markers of recurrence and “Pan-cancer” biomarkers. Titles and abstracts identified by electronic search were reviewed independently by two reviewers, using Rayyan. Conflict was resolved by a third independent reviewer. Results The electronic search yielded 2619 records. After removal of duplicates 2267 records remained for screening. Following screening of titles and abstracts, 130 studies remained for full-text review. After full-text review, 53 records fulfilled the criteria for detailed analysis and reporting. 86 additional papers were identified through manual searches and review of reference lists. Studies were published from diverse populations including: USA, China, Japan, Germany, Poland, India, Australia, Ireland, South Africa, Italy, Brazil, Turkey and Netherlands. Blood biomarkers investigated in the literature comprised classical tumour markers, such as CEA, SCC-AG, CA19–9, CA-125; inflammatory markers; metabolites; peptides and glycoproteins; enzymes; cytokines; volatile organic compounds; circulating tumour cells; miRNA; circulating DNA; long non-coding RNAs; circular RNAs; peripheral blood cell DNA/RNA; antibodies and tumour-associated antigens; and spectra from Raman spectroscopy. The most widely described markers were antibodies and tumour-associated antigens (61 papers) and miRNAs (25 papers). Several other panels with combinations of different types of markers are described. The highest area under receiver operating curve (AUC) was 0.99 from two papers. One paper described a panel of serum miRNA ratios (RNU6-1/miR-16-5p, miR-25-3p/miR-320a, let-7e-5p/miR-15b-5p, miR-30a-5p/miR-324-5p, miR-17-5p/miR-194-5p). The other described a serum N-glycan with mass:charge ratio of 2244. Conclusions Several blood biomarkers with a potential role in screening or diagnosis of oesophageal cancer have been described in the literature. The results of some tests suggest promising diagnostic accuracy with two papers having AUCs of 0.99. Despite this however, no single test has come into clinical practice as a non-invasive diagnostic test. The gold standard in diagnosing oesophageal squamous cell carcinoma and oesophageal adenocarcinoma remains to be endoscopy and biopsy. Reasons for these markers not being utilised in clinical practice to date include need for further studies to validate results with larger sample size; challenges in reproducibility of studies; cost and lack of availability of techniques used in hospital; and other challenges to translate scientific studies into clinical application. There seem to be promising candidates liquid biopsy markers in oesophageal cancer, but further work is required to validate and translate the use of some promising markers into clinical practice.