AbstractObjective:To determine the temporal directionality of the association between food insecurity and maternal depression.Design:Food insecurity was measured at two time points using the eighteen-item USDA Food Security Scale. Maternal depression was measured at two time points using the fifteen-item Composite International Diagnostic Interview-Short Form. Two structural equation models were utilised to evaluate the impact of food insecurity on maternal depression (model 1) and the impact of maternal depression on food insecurity (model 2). Both models controlled for socio-demographic and parenting characteristics and child behaviour problems, along with prior measures of the dependent variable and concurrent measures of the independent variable.Setting:Fragile Families and Child Wellbeing (FFCW) study, twenty cities across the USA.Participants:4897 mothers who participated in two waves of the FFCW study.Results:On average, 17 % (time 1) and 15 % (time 2) of mothers experienced food insecurity and 21 % (time 1) and 17 % (time 2) of mothers experienced depression over time. Maternal depression at time 1 was associated with 53 % increased odds (OR = 1·53; B = 0·43;P< 0·001) of food insecurity at time 2, controlling for time 1 food insecurity, concurrent depression and covariates. Food insecurity at time 1 was associated with 36 % increased odds (OR = 1·36; B = 0·31;P< 0·001) of maternal depression at time 2, controlling for time 1 depression, concurrent food insecurity and covariates.Conclusions:We found a bidirectional relationship between food insecurity and maternal depression. A holistic approach that combines food assistance and mental health services may be an efficacious approach to reducing both depressive symptoms and food insecurity among low-income mothers.