A previous report suggested that 88 of individuals in the general population with total cholesterol (TC) > 9.3 mmol/L have familial hypercholesterolaemia (FH). We tested this hypothesis in a cohort of 4896 {UK} civil servants, mean (SD) age 44 (±6) years, using next generation sequencing to achieve a comprehensive genetic diagnosis. 25 (0.5) participants (mean age 49.2 years) had baseline TC > 9.3 mmol/L, and overall we found an FH-causing mutation in the {LDLR} gene in seven (28) subjects. The detection rate increased to 39 by excluding eight participants with triglyceride levels over 2.3 mmol/L, and reached 75 in those with TC > 10.4 mmol/L. By extrapolation, the detection rate would be ∼25 by including all participants with TC > 8.6 mmol/L (2.5 standard deviations from the mean). Based on the 1/500 {FH} frequency, 30 of all FH-cases in this cohort would be missed using the 9.3 mmol/L cut-off. Given that an overall detection rate of 25 is considered economically acceptable, these data suggest that a diagnostic {TC} cut-off of 8.6 mmol/L, rather than 9.3 mmol/L would be clinically useful for {FH} in the general population.