Job loss is known to adversely affect mental health, but the time course of recovery and the role of anticipation remain unclear. Using 22 annual waves (2001-2022) of the Household, Income and Labour Dynamics in Australia (HILDA) survey, we estimate fixed-effects models to examine the relationship between redundancy and mental health (SF-36), incorporating subjective probability of job loss to refine anticipation measures. The final sample consists of 14,195 individuals and 4,251 redundancy events. Three key findings emerge. First, we document a generalized decline in mental health prior to job loss that is not confined to individuals who anticipate redundancy, suggesting psychological costs of impending job loss due to factors other than anticipation. Second, we document complete recovery among those who are re-employed, revealing that psychological restoration can occur relatively quickly upon securing new employment. Third, perceived anticipation of job loss does not appear to meaningfully alter these post-redundancy recovery trajectories. These findings call for greater emphasis on employment trajectories in both research and policy aimed at understanding and mitigating the mental health impacts of job loss.