Methods: 145 patients, who had received radiotherapy for cervical carcinoma more than three years ago, were invited for an assessment to answer questionnaires with a research nurse and offered a medical consultation. The questionnaires were devised from the published scores for LENT SOMA to obtain subjective scores of late effects at each organ site affected by the pelvic radiotherapy fields. This work was the result of collaboration with the departments of clinical oncology, psychological medicine and medical statistics. The questionnaires evolved over a period of 18 months with the fourth draft deemed satisfactory. Questions were used to obtain the objective evidence of treatment effect along with clinical examination and review of the medical notes to assess and record management and analytic evidence based on imaging results. Quality of life was assessed using the EORTC Core30 questionnaire (with permission) and scored according to the algorithm provided by the EORTC QOL study group (Aarondsen et a1,1993). The Hospital Anxiety and Depression Scale questionnaire was also used. Results: 100 patients attended for interview and assessment. Scores for each organ site - 6 sites uterus/cervix, ovary/reproductive, rectum/bowel, bladder/urethra, ureter/kidney and vagina were obtained. Total scores were calculated for subjective and objective criteria using an average score as recommended by the working groups that devised the LS scales. In addition, the maximum scores for each organ site were obtained. The objective LS scores for rectum/bowel correlated significantly with the FI scores, rho =0.79,p<0.0005 but there was less correlation with subjective scores, rho=0.32, p=<0.001. The objective LS score for bladder/urethra was significantly correlated with the FI score, rho =0.71 ,p<0.0005 but the subjective score was weakly correlated, rho=0.43,p<0.0005 (see figure). The total LS objective score was correlated with the total FI score, rho=0.62, p<0.0005. The total LS objective score and maximum FI score was correlated, rho=0.57, p<0.0005. The maximum scores and mean scores for each of the subsites was compared with the FI scores. Using the mean scores gave a slightly greater range and may allow greater discrimination than when using the maximum LENT SOMA score for a subsite. The total EORTC score was not well correlated with the LS objective criteria rho = -0.32, p=0.001, though there was better correlation with the subjective LS scores rho= -0.55, p<0.0005. Similarly, there was better correlation with the HAD anxiety score with the LS subjective criteria rho=0.52,p<0.0005 and HAD depression rho=0.48,p<0.0005, but the objective LS criteria were less well correlated for the HAD anxiety score, rho=0.33,p=0.001 and HAD depression, rho=0.27, p=0.007. Conclusions: The objective LENT SOMA scores were well correlated with the maximum FI score for organ sites, but less so with a total LS score. The data suggest that it may be better to look at the scores for each subsite rather than a total LS score. The subjective LS scores were less well correlated with the FI scores and this Js not surprising as the FI scoring is based on objective scoring and not taking the patients' views into account. The subjective LS scores were correlated with the EORTC and HADS scores but the objective LENT SOMA score was poorly correlated with the quality of life measurements.