A major challenge of wildlife telemetry is choosing an attachment technique that maximizes transmitter retention while minimizing negative side effects. For waterbirds, attachment of transmitters with subcutaneous anchors has been an effective and well-established technique, having been used on >40 species. This method was recently modified to include a second subcutaneous anchor, presumably increasing transmitter retention beyond that of single-anchor attachments. This putative benefit may be offset, however, by increased health risks related to additional incisions and subcutaneous protrusions. To test this potential trade-off, we attached radiotransmitters to molting and wintering surf (Melanitta perspicillata) and white-winged scoters (M. fusca) during 2008 and 2009 in Washington State and southeast Alaska, USA, using single- (121 scoters) and double-anchor (128 scoters) attachment techniques. We estimated daily probabilities of survival and radio retention for each group, this being apparent retention for wintering scoters because we could not differentiate shed transmitters from flighted emigration. For scoters during the flightless remigial molt, we found that addition of a second anchor increased cumulative retention probability (±SE) over a 49-day period from 0.69 ± 0.11 for single-anchor to 0.88 ± 0.07 for double-anchor attachments, while having no effect on survival. However, during winter, scoters with double-anchor attachments experienced no improvement in apparent retention, while having significantly lower survival during their first 14 days following transmitter attachment; of 15 mortalities during this period, 11 had 2 subcutaneous anchors. From day 15 onward, winter survival rates were nearly identical for single- versus double-anchor attachments, indicating that adverse effects of subcutaneous anchors were mainly limited to the 14-day postattachment period. Overall, given that the survival cost of adding a second subcutaneous anchor was substantial for wintering scoters—decreasing 14-day survival by 12% for adults and 23% for juveniles—we recommend that researchers opt for single-anchor attachments under most circumstances, especially during winter when birds may be energetically challenged.