A focus in the literature on determinants of women's health is the cost and benefit of occupying multiple roles as employee, spouse, and mother, yet little attention has been given to the work and home characteristics of different roles for women in paid and unpaid work. The impact of work-home factors on socio-economic gradients in health has also tended to be overlooked. This paper assesses the contribution of work-home factors on socio-economic differences in psychological distress among women, using data from the 1958 British birth cohort. Outcome measures include psychological distress and social class at age 33. Work-home measures include: (1) roles--employment, marital status, domestic responsibility and parental status (2) work characteristics--psychosocial job strain, insecurity, unsocial working hours, and (3) home characteristics youngest child's age, total number of children, childcare responsibilities and having an older adult in the household (over 70 years). A social gradient in psychological distress exists: the odds ratio (OR) for classes IV and V versus. I and II was 3.02, adjusting for prior psychological distress reduces this to 2.36. Whilst, work and home factors were associated separately with distress and social class, the combined effect of work and home factors did not account for the class gradient in distress. This surprising result therefore implicates factors beyond adult social roles examined here in the development of socio-economic gradients.