Many workers severely exposed to inhaled particles have remained unaffected--what has protected them? Some workers have been affected despite mild exposure--what has rendered them vulnerable? There has been little progress in answering even the first question, and there can be no guarantee that factors rendering the lightly exposed vulnerable are the converse of those protecting the heavily exposed. An answer to the second question would be useful, say, for screening. But it requires study at the low end of the exposure scale, and there are many variables, probably intricately related, to be taken into account. A case/referent design might seem an obvious approach, but in its classical form it incurs essential and perhaps insurmountable difficulties. A longitudinal epidemiologic study would have to be very large, very complex, and very long in duration. Animal experiments also would have to be large and complex, probably embracing most of the animals' life span, but even so they could not take all relevant variables into account. It is recognized that the present paper has emphasized many difficulties and may thus appear negative in approach. Furthermore, although a compromise epidemiologic design is put forward, it is speculative.