CGAP convened a group of industry experts to look at client over-indebtedness issues. What exactly do we mean by over-indebtedness? How can we tell how often it's occurring? Are long-term default rates a good proxy? What does existing research tell us about it? These questions proved to be surprisingly thorny. Evidence from Bolivia and Bangladesh indicates that borrowers sometimes have to take drastic steps to repay their loans, but these studies shed little light on the question of whether these borrowers would have been better off without the loan. A tentative conclusion is that we may need some expensive randomized impact studies in order to determine whether easily-measured variables like repayment rates, or percentage of disposable income devoted to debt service, or clients' own statements about whether they feel debt-stressed, are useful proxies for the likelihood that a loan is hurting a borrower.