Objectives: The aim of this study was to examine the variation of enterolactone from fasting and non-fasting blood of middleaged healthy women eating a normal diet to determine the usefulness of a single sample in epidemiological studies. Subjects and methods: Twenty-six women born between 1940 and 1950 were recruited within the Malmö Diet and Cancer cohort. Three non-fasting and two overnight fasting samples were collected from each individual during a 5-week period. Twenty-one participated in all measurements. Enterolactone concentrations were analyzed by time-resolved fluoroimmunoassay. Results: The within-subject and between-subject variations (coefficient of variations, CV) were estimated to 59 and 89% respectively for fasting samples and 71 and 67% for non-fasting samples. The intraclass correlation coefficients (ICC) were estimated to 0.66 (95% confidence interval (CI) 0.35-0.84) for fasting and 0.48 (95% CI, 0.22-0.72) for non-fasting samples. Conclusions: Although the estimated ICC for blood samples was moderate, it indicates that enterolactone levels of both fasting and non-fasting blood samples should be useful in future projects within the Malmö Diet and Cancer cohort.