Background The Dutch Measure for Quantification of Treatment Resistance in Depression (DM-TRD) is a promising prediction tool for major depressive disorder (MDD) based on variables associated with treatment outcome. The objective of our study was to examine the association between the DM-TRD and clinical course in a large cohort of MDD outpatients receiving treatment as usual. Furthermore, we examined whether the addition of an item measuring the presence of childhood adversity improved this association. Methods We included 1115 subjects with MDD (according to the DSM-IV) who were naturalistically treated at seven outpatient departments of a secondary mental healthcare center in the Netherlands. Data on subjects who had a diagnostic work-up between June 2014 and June 2016 were analyzed. Multilevel analyses were performed to examine the association between the DM-TRD score at baseline and clinical course, defined by symptom severity according to scores on the Quick Inventory of Depressive Symptomatology-Self Report (QIDS-SR) over time. We also investigated whether an extra item measuring childhood adversity improved the model. Results The model including the DM-TRD and its interaction with time was superior to previous models. The addition of childhood adversity and its interaction with time did not improve the model. Conclusions In depressed outpatients receiving treatment as usual, the solid longer-term association between higher DM-TRD scores and worse clinical course supports its usefulness in clinical practice. Childhood adversity did not improve the model value indicating that-counterintuitively-this parameter offers no additional predictive power to the variables included.