We quantified associations between IgE and cancer in a meta-analysis and cohort study. Pubmed and Embase were searched to extract information using predefined inclusion criteria. In the Apolipoprotein MOrtality RISk (AMORIS) database, 24,820 persons had IgE measurements. Multivariate Cox proportional hazard models were used to analyze associations between IgE and cancer. Twenty-seven studies were reviewed from which seven case-control studies were included for analysis. The pooled relative risk (random effects model) was 0.97 (95% CI 0.86-1.09). Cell types of tumor origin (mesenchymal tissue or cells of the nervous system, lymphatic or hematopoietic tissue, and epithelium) modified the effect. In the AMORIS cohort, 862 persons developed cancer. Hazard ratios comparing quartiles of IgE were similar to the findings in the meta-analysis (HR 0.87 (95% CI 0.72-1.06); 0.94 (0.78-1.14); 0.90 (0.74-1.10) for the 2nd, 3rd, and 4th quartile compared to the 1st quartile), but there was no pattern by tumor origin. Both studies showed a weak inverse association between IgE and cancer, but a pattern by cancer type was only seen in the meta-analysis. Our findings suggest the need for prospective studies studying IgE and cancer. Measurements of IgE should be combined with other information, e.g., bio-banked samples containing other key immunological discriminators.