This study evaluated the impact of Medicare's new drug benefit (Part D) on utilization of psychotropic medications and the consequent financial burden for the elderly population. The effect of Medicare Part D was measured by analyzing 24-month pharmacy claims collected from one of the largest retail pharmacy chains in the United States. Segmented regression analysis of interrupted time series was used to evaluate population-level changes in the utilization of the most commonly used psychotropic therapeutic categories among seniors, namely antidepressants, antipsychotics, and benzodiazepines. In the first-year implementation of Part D, the proportion of out-of-pocket payment in total pharmacy reimbursement decreased 18% for antidepressants (net savings of $4.52 per prescription) and 21% for antipsychotics (net savings of $5.71 per prescription). In contrast, the out-of-pocket share paid for benzodiazepines increased 19% (net increase of $2.79 per prescription). Part D was associated with significant month-to-month increase in use of antidepressants and antipsychotics. By December 2006, the antidepressant and antipsychotic prescriptions filled by seniors grew 7% (from 273,166 to 293,590 prescriptions per month, p<.001) and 18% (from 41,079 to 48,276 prescriptions per month, p<.001), respectively, compared with the expected level estimated from prior Part D trends. In contrast, Part D led to an immediate and sustained drop of 5% (from 238,961 to 226,622 prescriptions per month, p<.001) in number of benzodiazepine prescriptions filled by seniors. Part D has improved access to psychotropic medications covered under plans by reducing out-of-pocket expenses. However, the financial burden related to excluded medications, such as benzodiazepines, has significantly increased.