Imaging is playing an increasing role in all stages of radiotherapy, with images from different modalities being interpreted in the same clinical and geometrical contexts. However, the methods used to assess image quality have historically differed between modalities and are often subjective, hindering inter-modality comparisons and overall optimisation. The ‘IQ Works’ package aims to overcome this by performing a fully objective, quantitative and consistent performance evaluation of any radiotherapy imaging modality. Regardless of the specific modality, most phantoms for image quality evaluation tend to consist of the same atomic test components in different geometrical configurations. An extensible framework has been developed comprising modules for analysing each of these components. ‘Analysis trees’ for particular phantoms are interactively constructed from these modular building blocks, then stored in a phantom library. These can be recalled for subsequent rapid and fully-automatic processing of any phantom's images. Analysis modules include the assessment of alignment, geometric linearity, characteristic curve (including Hounsfield Unit/electron density calibration), uniformity, spatial resolution in terms of modulation transfer function (MTF), noise power spectrum (NPS) and detective quantum efficiency. Each module is rigorously tested on standard test images and compared against independent methods. Comprehensive reports are generated and the results stored in a database to facilitate trend and statistical analyses. Individual test results are checked to ensure they lie within pre-defined tolerances, with the user being alerted if not. For five years, IQ Works has been utilised routinely in the quality assurance program in Edinburgh. The new system is more comprehensive and robust, yet also far more straightforward and efficient, reducing the time required for routine testing to less than a fifth. As such, complex assessments are now routinely performed by therapists rather than physicists. All analysis modules produce results in close agreement (<1%) with independent methods. Analysis schemes have been developed for multi-slice and cone-beam CT, digital radiography/fluoroscopy (simulators and linac-mounted), EPIDs, DRRs, MRI and soft-copy displays. Detailed analyses have yielded valuable results, such as highlighting differences of >6% in CT HU calibration curve depending on acquisition settings and providing early indicators of deterioration requiring technical intervention. IQ Works has assisted in commissioning linac-mounted cone-beam CT by enabling clinically relevant comparisons with conventional CT scanners. The IQ Works system can readily model new commercial and bespoke phantoms and is an efficient means of objectively assessing the image quality of radiotherapy imaging modalities. For the first time, meaningful inter-modality comparisons of image quality throughout the treatment chain can be performed, leading to better overall optimisation. Furthermore, complex assessments can be performed by any staff group. The package has been adopted as the basis of a new open-source analysis system launched by a consortium consisting of the IPEM, CT Users Group and other parties in the UK. The package may be downloaded free from http://www.iqworks.org.