PURPOSE Accurate delineation of target volume and normal tissue is critical for intensity modulated radiation therapy (IMRT) use in cervical cancer. Phase III Multicentre Trial of Weekly Induction Chemotherapy Followed by Standard Chemoradiation versus Standard Chemoradiation Alone in Patients With Locally Advanced Cervical Cancer (INTERLACE) radiation therapy quality assurance (RTQA) has highlighted significant interobserver delineation variation. Prescriptive guidelines reduce interobserver variation in other cancers. METHODS AND MATERIALS A literature search using PubMed/Medline database of guidelines for target anatomy delineation in cervical cancer was undertaken. Differences in practice in these publications and INTERLACE trial RTQA were identified. Consensus best practice delineation was derived and a pictorial atlas produced. The proportion of outlines complying with protocol in test and real-time cases was compared before and after atlas implementation within the INTERLACE RTQA pack. RESULTS Seven key papers were reviewed. Eleven areas of variation were identified. These included the definition and editing of bowel, definition of the femur, vagina, parametria, inferior and superior nodal borders, nodal clinical target volume (CTV) editing, para-aortic nodal CTV definition, and the margin to be used around enlarged nodes. The average proportion of outlines (of 4; primary CTV, nodal CTV, bladder, rectum) complying with protocol in test and real-time cases improved from 1.8 to 2.7 (difference, 0.9; 95% confidence interval, 0.3-1.5; P = .003) with atlas use. CONCLUSION Differences exist in the published literature and clinical practice. This pictorial atlas reflects consensus recommendations and is now available to INTERLACE participating centers. Atlas use has reduced interobserver delineation variation in this trial setting.