A moral hazard problem was investigated by analysing the individual behaviour of female and male employees with regard to utilisation of sickness insurance in connection with perceived job security. It was hypothesised that employees with a higher perceived job security take more frequent sickness absence. Perceived higher job security is indicated by three variables, namely a permanent job contract, no unemployment history, and native ethnicity. The effect of perceived job security is expected to be stronger on short-term than on long-term sickness absence, since a medical certificate is required for the latter. Public health survey data from Stockholm County, Sweden, covering the year 2002 was used. Using logistic regression analyses separately for short- and long-term sickness absence and for females and males, we found that short-term sickness absence is more strongly influenced by perceived job security than long-term sickness absence. We observe indications of moral hazard in both female and male employees. However, the three indicators of perceived job security have a different influence on females and males.