We estimated electrocution rates for raptors and common ravens (Corvus corax) for the Moon Lake Electrical Association in northeastern Utah and northwestern Colorado, USA. From July 2001 to May 2003, we conducted mortality searches at randomly selected distribution line segments and poles within 3 regions, but rate estimates (0.0036–0.0112 deaths/pole/yr) may have been biased by the effects of scavengers and by long sampling intervals (≥3 months), which prevented us from determining the cause of death for most birds because of advanced decay. In 2002–2003, we conducted carcass removal experiments in the Rangely Oil Field (ROF) in northwestern Colorado to estimate scavenging effects, and in 2003–2004, we reduced sampling intervals to 1 month and searched for dead birds at all distribution poles in the ROF. The shorter sampling interval nearly tripled the number of birds suitable for necropsy, but we were still unable to establish cause of death for >40% of our sample. Instead of eliminating the unknowns from rate estimates, we estimated minimum annual electrocution rates using only confirmed electrocutions and maximum annual electrocution rates based on all available mortalities, including mortalities without known causes. Golden eagles (Aquila chrysaetos) accounted for 63% of dead birds found in 2003–2004, but they were removed by scavengers at rates well below hawks and owls (6.8% vs. 55.6%). We compared maximum rates for the ROF in 2003–2004 with the rates estimated from a survey conducted at the same poles in 1999 to assess the effects of retrofitting conducted by Moon Lake from 1999 to 2003. Electrocution rates in 2003–2004 were 47% lower than those in 1999. Raptor densities in the ROF did not change during our study, suggesting the reduction was not the result of changes in raptor populations. However, estimates of raptor densities in 1999 were not available, and we cannot be sure that numbers of birds using the oil field in 1999 were similar to those in 2003–2004. Our research emphasizes the difficulties of estimating electrocution rates precisely but suggests that utilities will have the greatest effect on mortality by monitoring power lines at large scales and focusing subsequent mitigation efforts in areas that pose the greatest risk to the greatest number of birds.