OBJECTIVES: We report on differences in medication adherence for patients under routine care newly initiated on an atypical antipsychotic (“rookies”) and for patients continuing treatment (“veterans”). The same analysis was performed for patients who received antidepressants. We hypothesized that newly initiated patients might face a greater risk of discontinuation due to illness or medication inexperience. METHODS: Deidentified pharmacy data from throughout the United States was used. Rookies were patients without a history of antipsychotic or antidepressant use in the 180 day period prior to their first fill. Veterans were those who continued a previously initiated antipsychotic or antidepressant. The duration of medication use was determined by using Kaplan-Meier time to discontinuation analysis. Discontinuation was defined as being 30 days late for a scheduled refill. The antipsychotic sample (N= 406,032) included risperidone (n=98,257), olanzapine (n= 84,919), quetiapine (n=78,873), aripiprazole (n=20,279), ziprasidone (n=121,607), or clozapine (n=2,097). The antidepressant sample (N=211,565) included fluoxetine (n=43,412), sertraline (n=58,388), paroxetine cr (n=20,296), venlafaxine xr (n=28,983), citalopram (n=22,904), or escitalopram (n=37,582). RESULTS: The median time to discontinuation for atypical antipsychotic rookies was 34 days and 125 days for veterans. The median time to discontinuation for antidepressant rookies was 67 days and 184 days for antipsychotic veterans. CONCLUSION: Patients without atypical antipsychotic or antidepressant use in the preceding 180 days appear to face an especially high risk of medication discontinuation in the first 30-60 days after starting medication and merit close follow-up at this time. Such data suggests that what patients know and do as a result of prior medication or illness experience (or lack of it) contributes in important ways to their subsequent likelihood of continuing medication.