Human epididymis 4 (HE4) is a promising ovarian cancer biomarker, but it has not been evaluated in primary care. In this prospective observational study, we investigated the diagnostic accuracy of HE4 alone and in combination with CA125 for the detection of ovarian cancer in symptomatic women attending primary care. General practitioner (GP)-requested CA125 samples were tested for HE4 at a large teaching hospital in Manchester, and cancer outcomes were tracked for 12 months. We found a low incidence of ovarian cancer in primary care; thus, the cohort was enriched with pre-surgical samples from 81 ovarian cancer patients. The Risk of Ovarian Malignancy Algorithm (ROMA) was calculated using age (</>51) as a surrogate for menopause. Conventional diagnostic accuracy metrics were determined. A total of 1229 patients were included; 82 had ovarian cancer. Overall, ROMA performed best (AUC-0.96 (95%CI: 0.94-0.98, p = <0.001)). In women under 50 years, the combination of CA125 and HE4 (either marker positive) was superior (sensitivity: 100% (95%CI: 81.5-100.0), specificity: 80.1% (95%CI 76.7-83.1)). In women over 50, ROMA performed best (sensitivity: 84.4% (95%CI: 73.1-92.2), specificity: 87.2% (95%CI 84.1-90)). HE4 and ROMA may improve ovarian cancer detection in primary care, particularly for women under 50 years, in whom diagnosis is challenging. Validation in a larger primary care cohort is required.