Inflammatory bowel disease (IBD) is typically diagnosed after the onset of symptoms in the context of established, characteristic patterns of intestinal inflammation. However, there is now substantial evidence pointing to a prolonged, biologically active preclinical phase of disease. Analysis of archived biological samples from large-scale longitudinal cohort studies of healthy individuals, some of whom develop incident IBD, has identified different molecular features that can be detected many years before clinical presentation. These include increased titres of antimicrobial and autoreactive antibodies and perturbations in a complex network of circulating, immunologically active proteins. As well as affording ‘diagnostic’ opportunities to identify individuals destined to develop IBD, an integrated view of these multiple different molecular features enables speculation of potential proximal drivers of preclinical IBD. Consistently recognised associations include dysregulated mononuclear phagocyte–lymphocyte interactions, augmented chemotaxis, frequently relating to interferon-–driven chemokine programmes and evidence of early tissue injury, such as increased circulating extracellular matrix components and metalloproteinases. Increased levels of circulating antibacterial and antiviral antibody responses hint towards disordered host–microbe interactions as potential prime triggers for the transition between health and early disease, although it is possible that these serological responses are an epiphenomenon linked to early mucosal damage and microbial translocation. There is now a timely opportunity to develop these different molecular features into scalable and clinically tractable biomarker panels to detect preclinical disease and enable strategies to proactively intercept IBD before it even develops.