It is widely assumed that contingent forms of employment, such as fixed-term contracts, labour-hire and casual employment, are associated with low quality jobs. This hypothesis is tested using data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey, a nationally representative household panel survey covering a country with a high incidence of non-standard employment. Ordered logit regression models of job satisfaction are estimated that hold constant all time-invariant individual differences as well as a range of observed time-varying characteristics. The results indicate that, among males, both casual employees and labour-hire workers (but not fixed-term contract workers) report noticeably lower levels of job satisfaction. Restricting the sample to persons aged 20-59 increases the estimated magnitudes of these effects. Negative effects for women are mainly restricted to labour-hire workers. We also show that the relationships between job satisfaction and contract type vary with educational attainment and the length of job tenure. Working hours arrangements also mediate the relationship.