Longitudinal survey data describing 6,190 subjects are analyzed using log-linear methods to determine which, if any, of three hypothesized links between short-term community economic change and illness or injury is correct. Assumptions of these three possible links are, respectively: (1) Economic contraction increases the incidence of undesirable job and financial events that, in turn, increase the incidence of illness and injury; (2) Economic change per se increases the incidence of undesirable job and financial events and, therefore, of illness and injury; and (3) Economic change per se increases the incidence of all job and financial events and, therefore, of illness and injury. The data support the first hypothesized connection, but the process is observed only in middle SES respondents.