The British Institute of Radiology was the venue in May 2004 for an excellent conference addressing issues regarding the recording and analysis of late effects of radiotherapy treatments. Currently there is increasing interest in radiotherapy dose escalation using conformal treatments and novel therapies such as intensity-modulated radiotherapy. In view of the steep dose–response relationship with radiotherapy, dose escalation can be dangerous unless undertaken carefully with accurate reporting of late serious adverse effects, so that the therapeutic ratio of treatments can be determined. The meeting attempted to address issues around assessment of treatment outcomes with a view to consensus regarding current and future practice. Professor Anne Barrett outlined some of the work in this area achieved by the Recording, Education and Amelioration of the Consequences of Treatment group (REACT). REACT was set up under the auspices of the European ESQUIRE Project [1]. ESQUIRE is co-ordinated by the European Society for Therapeutic Radiology and Oncology with the objective of improving treatment outcomes for cancer patients by addressing various facets of radiotherapy implementation and training. Currently there are a variety of systems in place for the recording of late effects of treatment, including from the RTOG/ EORTC, NCI-CTC3 (Common Toxicity Criteria) and the Anglo-Franco system, in addition to inhouse nonvalidated approaches. Professor Barrett also presented results of an audit undertaken of follow-up practice in Europe. Four centres in the UK were involved in this study. In over 50% of visits patients and doctors reported disease or treatment related problems. There was an attribution bias in patients who, compared with doctors, were more likely to consider their problems related to disease than treatment. The use of a short scale to assess late morbidity of treatment, developed by Professors Dische and Saunders (Table 1), increased the recording of side effects from 23% before the study, to 90% [2]. Grade 3 or 4 late morbidity was reported to be in the range of 4–9%. Detailed recording of late effects is difficult, given the complexity and time pressures existing in clinics. The LENT SOMA system is one of the most common systems for recording late effects [3]. It is recognized as being an extremely useful tool, as it includes both subjective as well as objective scoring measures, but is cumbersome and difficult to use in daily practice. Dr Susan Davidson from the Christie Hospital, Manchester, discussed their experience of the use of the LENT SOMA system for analysing late effects of radiotherapy treatment. There remain problems in defining what ‘‘acceptable toxicity’’ is, and this varies depending on whether it is a medical or patient view. Validated systems for assessing late effects need to be sensitive, reproducible, feasible and convenient. Although the LENT SOMA system has many of these features, it still could be improved further. The Christie Hospital has developed patient questionnaires based on the LENT SOMA for certain tumour sites, and suggest these could be developed into more useful tools. Details of these questionnaires can be found at: www.christie.nhs.uk/ proinfo/departments/clinical_oncology/lent_soma.htm What was clear is that the use of different reporting systems across the UK, Europe and worldwide makes cross comparisons of differing treatments difficult. This lack of a standardized approach was a theme throughout the conference, as was the need to simplify data collection and educate regarding the importance of recording late effects. Dr Catharine West, from the Christie Hospital, tried to address the question of whether it is possible to predict normal tissue effects of radiotherapy in advance of treatment. She provided an excellent historical overview of the development of radiobiology understanding through the 20th century to the modern day. Recent developments make the possibility of predictive assays tangible, although as yet not practicable. In the post genomic