A quasi-experimental field investigation was conducted with blast furnace workers from two German steel companies. Of the 180 workers participating in the study 123 blast furnace workers at comparable work places were taken into the analysis. 48 blast furnace workers of a company offering only highly insecure employment ("Experimental Group", EG) and 75 from a company with no job insecurity ("Control Group", CG) were compared by means of perceived job insecurity, three aspects of well-being and strain (job satisfaction, irritation/strain, psychosomatic complaints), five facets of social support, and five facets of control at work. First, statistical comparisons confirm the strong difference in job security between EG and CG by means of perceived job insecurity and though the validity of the independent variable. Moreover, for the EG comparisons show no substantive increase in psychosomatic complaints or job strain but diminished job satisfaction, they also indicate reduced collective control, hope for control, and reduced perceived alternatives on the labour market, less support from supervisors and more support from friends. Second, the results were differentiated by moderated regression analysis with respect to the five facets of social support and five components of control at work. Results indicate moderation effects and several main (additive) effects for the social support and control at work variables, however the moderation effects are not homogeneously distributed among the three dependent variables and in addition their interpretation as buffering or coping effects is ambiguous. In total, control at work turns out to be the most influencing resource with respect to the stress from job insecurity. The results are discussed for example with respect to the role of resources in the unemployment process and with regard to several methodological questions like the advantage of objective measures of job insecurity or a theory-driven differentiation of moderator variables and moderator assumptions.