We investigate the reliability of the 'furfine filter' often used to identify interbank loans and interest rates from interbank payments settled at central banks. To this end, we have been granted access to records of all unsecured overnight interbank loans during a month from the banks that participated in norges bank's real-time gross settlement system. The filter applied was able to identify each of these loans and correctly derive the associated interest rates. The filter's reliability is also supported by additional evidence based on the norwegian overnight weighted average (nowa) interest rates beyond the survey month. Sensitivity analyses suggest the share of false or overlooked loans may remain small if the filter design largely incorporates interbank market conventions regarding loan size requests and interest rate quotes.