Background Every year, the Health Bus (Bus Sante) enables a representative sample of around 1000 members of the non-institutionalised adult population of Geneva, Switzerland, to undergo a health evaluation including physical measurements, blood tests and questionnaires. A published study using 1999-2009 Health Bus data found that body-mass index (BMI) was not associated with a standard measure of sedentariness: 10% or less of total daily energy expenditure spent in activities demanding at least 4 metabolic equivalents. We wanted to know if this remained true using more recent data, if daily walking was separated from other physical activities, or if other subclinical endpoints were considered. Methods Results for 1102 women and 1000 men aged 20-80 who attended the Health Bus in 2013-2014 were obtained. We combined three items from the Physical Activity Frequency Questionnaire – minutes per week spent walking normally, fast/uphill, and while carrying/pulling a load – to create a metric for daily walking (combining walking for transport and for leisure). We compared this metric to results obtained by the participants for 8 subclinical endpoints, including BMI. We fitted a generalised multivariate linear model on SPSS. The significance of associations between variables was evaluated using the F-statistic (with p Results We found that sedentariness was not significantly associated with BMI. Our combined walking metric was not associated with BMI, but was with resting heart rate. When the model was corrected for age, gender, income and level of education, we found significant associations with all investigated endpoints except BMI. This corrected model included: waist-to-hip ratio, total glycaemia, total cholesterol, total triglycerides, blood pressure (systolic and diastolic) and resting heart rate. Considered separately in the corrected model, walking normally was associated with none of the endpoints. Rapid/uphill walking was associated only with waist-to-hip ratio and blood pressure (systolic and diastolic). Walking while carrying/pulling a load was associated only with resting heart rate. Conclusions The results of this retrospective study suggest that daily walking is not related to BMI in this population-based sample, but that it may be related to other relevant subclinical endpoints. Separating walking into three categories and health effects into eight endpoints may provide a useful framework for future research. Limitations: The subclinical endpoints were measured in a standardised way on a representative sample, whereas the walking depends on self-declaration. We suggest recruiting a subset of participants to a follow-on study to evaluate actual walking behaviour using pedometers and GPS technology.