OBJECTIVE: This study aims to compare the accuracy of digital mammography (DM), digital breast tomosynthesis (DBT), ultrasound (US), magnetic resonance imaging (MRI), and contrast-enhanced mammography (CEM) in the preoperative evaluation of breast cancer size in architectural distortions (ADs). Additionally, it assesses whether including thin spicules in mammography measurements affects accuracy. MATERIALS AND METHODS: We planned a retrospective analysis of invasive breast cancers presenting as ADs in our breast screening program between 2018 and 2022. Tumor size was measured in mm using DM, DBT, US, MRI, and CEM. Measurements were compared to the surgical specimen sizes. Two measurement approaches for DM and DBT were applied, considering and not considering thin spicules. T-student test was used to compare mean sizes across imaging techniques with the surgical specimen. RESULTS: The study encompassed 59 female patients with 63 ADs. Mean age was 60.1 years (Standard Deviation (SD): 6.3). The cancers included four histological subtypes, ductal (69.8%), lobular (23.8%), tubular (4.8%), and micropapillary (1.6%). All imaging techniques, except for US (mean: 12.4 mm, SD: 5.7), overestimated tumor size compared to histology (mean: 16.40 mm, SD: 9). CEM, MRI, and DBT without thin spicules closely matched histological size. Including thin spicules in DM and DBT led to overestimation. Concordance was highest with CEM (75%) and MRI (67.6%). No significant differences were found between ductal and lobular carcinoma. CONCLUSION: For preoperative tumor size estimation of breast cancer in ADs, DBT excluding thin spicules, CEM, and MRI seemed most accurate. Including thin spicules in mammography leads to overestimation. KEY POINTS: Question Identifying the most accurate imaging technique for preoperative tumor staging of architectural distortions (ADs) is crucial now that contrast-enhanced mammography (CEM) is widely implemented. Findings Measuring thin wispy spicules in ADs on digital (DM) and digital breast tomosynthesis (DBT) should be avoided, as they consistently overestimate pathological tumor stage. Clinical relevance Precise tumor size estimation in ADs is critical for proper staging and treatment planning. This study favors the use of DBT excluding thin spicules, CEM, and magnetic resonance imaging (MRI) for optimal accuracy.