To identify the risk factors associated with the development of skin tears in older persons four hundred and fifty three patients (151 cases and 302 controls) were enrolled in a case–control study in a 500-bed metropolitan tertiary hospital in Western Australia between December 2008 and June 2009. Case eligibility was defined by a skin tear on admission, which had occurred in the last 5 days; or, a skin tear developed during hospitalisation. For each case, two controls who did not have a skin tear and had been admitted within 1 day of the case, were also enrolled. Data collected from the nursing staff and inpatient medical records included characteristics known, or hypothesised, to be associated with increased vulnerability to skin tears. Data analysis included a series of multivariate stepwise regressions to identify a number of different potential explanatory models. The most parsimonious model for predicting skin tear development comprised six variables: ecchymosis (bruising); senile purpura; haematoma; evidence of a previously healed skin tear; oedema; and inability to reposition oneself independently. The ability of these six characteristics to predict who among older patients could subsequently develop a skin tear now needs to be determined by a prospective study.