Aims: To determine the variation in radiation therapy boost use in a nationwide study following adjustment of a national guideline in 2011, as well as to address the relationship to patient, tumour and radiation therapy institutional factors. Materials and methods: All invasive breast cancers and non-invasive breast cancers (ductal carcinoma in situ; DCIS) that received external whole-breast radiation between 2011 and 2016 were selected from the Netherlands Cancer Registry. Box plots were used to evaluate variation over time and logistic regression was carried out to address other factors influencing the variation. Funnel plots were constructed, with unadjusted and adjusted data for patient and tumour factors significantly affecting the use of a boost. Results: For breast cancer patients (n = 45,207), the proportion receiving a boost and its range decreased over the years from 37.3-92.7% in 2011 to 28.3-65.4% in 2016. This trend was not observed in DCIS patients (n = 6,844). Young age, large tumours, high grade and the absence of tumour-free resection margins were associated with boost use for both breast cancer and DCIS. For breast cancer, triple-negative tumour subtype and metastatic lymph node involvement were also associated with boost use. Institutional factors did not influence the use of a boost and institutional variation remained substantial after case-mix adjustments. Conclusion: Following adjustment of a nationwide implemented guideline, variation in radiation therapy boost use decreased in patients with breast cancer but not in patients with DCIS. Several tumour and patient characteristics were associated with boost use. Substantial institutional variation could not be explained by differences in patient, tumour or predefined institutional characteristics. (C) 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.