We wished to develop and validate a simple linear analogue self-assessment (LASA) scale to assess utility values in multicentre, multicultural breast cancer trials. We compared two variants of a LASA scale (score range 0-100) with different anchoring points [perfect health to worst possible health (subjective health estimation, SHE) and perfect health to death (SHED)] in 84 patients with advanced breast cancer. Feasibility was explored in the first 48 patients interviewed. Values from the LASA scales were compared with each other and with a time trade off (TTO) interview. Scores from the two LASA scales were highly correlated (r=0.8, P < 0.0001, Spearman). The relationship between TTO interview and SHE was confirmed with tests for trend across ordered groups (linear-trend test P < 0.001). Most patients preferred SHE to SHED. SHE scores (in which high scores indicate high-health-state values) were significantly different by type of treatment, time from diagnosis and age. Thus, significantly different mean SHE scores were obtained from patients receiving no treatment or hormone therapy, mild and intensive chemotherapy (ANOVA P=0.03) and from patients with diagnosis 2 years, 2-5 years, 5-10 years and more than 10 years before interview (ANOVA P=0.02). Older patients (> 56 years) had a lower mean on the SHE scale (53 vs 61; ANOVA P=0.04). We found that the two versions of the LASA scale were equivalent for clinical purposes. SHE appeared to provide a feasible, patient-preferred and valid alternative to lengthy TTO interviews in assessing the value patients attach to their present state of health in large-scale cancer clinical trials.