Background: The aim of this study was to determine the relationship between household income and measures of health at different points in the life course. Method: Analysis of second-wave data collected in 1990–1991 in a longitudinal study of three age cohorts (15, 35 and 55 years in 1987–1988) in the West of Scotland was performed. The subjects were 851 18 year olds, 801 39 year olds and 761 58 year olds on whom data on Income and a range of health measures were available (92% of those interviewed in 1990–1991). Results: Blood pressure was not significantly related to household income. Height, weight, waist:hip ratio, body mass index, pulse rate, long-standing illness and limiting long-standing illness showed linear associations with household income and FEV1/height2, number of recent malaise symptoms reported, number of checklist of 22 recent symptoms reported, GHQ scores and self-assessed health had curvillnear relationships with household Income. There were interactions between Income and sex for weight, waist:hip ratio, BMI and pulse rate and interactions between income and age for malaise, symptom checklist and self-assessed health. A threshold model of the relationship between income and health was not supported. Conclusion: The relationship of income to health varies not only according to the health measure chosen but by age and gender.