were asked whether they saw themselves as "employees", "foremen" or "managers").Compared to foremen, employees had a small and imprecisely estimated increased risk of all cause mortality (age adjusted hazard ratio 1.11 (0.96, 1.30)), whereas managers had a more marked decreased risk (age adjusted hazard ratio 0.75 (0.62, 0.90)). The strongest predictors of increased mortality were father's manual as opposed to non-manual occupation (fully adjusted hazard ratio 1.16 (1.03-1.31)); lack of car access (fully adjusted hazard ratio 1.12 (1.02-1.23)) and shorter stature, (an 2 indicator of material deprivation in childhood, fully adjusted hazard ratio 1.09 (1.02-1.16) per 10cm decrease). In the fully adjusted analyses, perceived work-place status was only weakly associated with mortality (hazard ratio for employees compared to foremen 1. 09 (0.93, 1.27), managers compared to foremen 0.91 (0.74, 1.10)). In this population it appears that objective material circumstances, particularly in early life, are a more important determinant of health than perceptions of relative status.Conversely, higher perceived stress was not associated with poorer health, presumably because, in this population, higher stress was not associated with material disadvantage. Together these findings suggest that, rather than targeting perceptions of disadvantage and associated negative emotions, interventions to reduce health inequalities should aim to reduce objective material disadvantage, particularly that experienced in early life.