Following the failure of Lehman Brothers in September 2008, short-term credit markets were severely disrupted. In response, the Federal Reserve implemented new and unconventional facilities to help restore liquidity. Many existing analyses of these interventions are confounded by identification problems because they rely on aggregate data. Two unique micro datasets allow us to exploit both time series and cross-sectional variation to evaluate one of the most unusual of these facilities-the Asset-Backed Commercial Paper Money Market Mutual Fund Liquidity Facility (AMLF). The AMLF extended collateralized loans to depository institutions that purchased asset-backed commercial paper (ABCP) from money market funds, helping these funds meet the heavy redemptions that followed Lehman's bankruptcy. The program, which lent $150 billion in its first 10 days of operation, was wound down with no credit losses to the Federal Reserve. Our findings indicate that the facility was effective as measured against its dual objectives: it helped stabilize asset outflows from money market mutual funds, and it improved liquidity in the ABCP market. Using a differences-in-differences approach we show that after the facility was implemented, money market fund outflows decreased more for those funds that held more eligible collateral. Similarly, we show that yields on AMLF-eligible ABCP decreased significantly relative to those on otherwise comparable AMLF-ineligible commercial paper.