The performance of three models of vibrating-wire strain gages is judged in light of laboratory investigation and field experience. Gage models include the Telemac SC-11, Telemac SC-8, and Geonor acoustical gages. Measurement accuracy is controlled by gage sensitivity to changes in strain, zero drift of the gage readings, and thermal response. The Telemac SC-8 and Geonor gages show temperature dependencies of 2.5 microstrain/10°F (4.5 microstrain/10 K) and 1.5 microstrain/10°F (2.7 microstrain/10 K), respectively. Both of these gage models show little or no zero drift with time. The Telemac SC-11 gage shows a temperature dependency of 4.6 microstrain/10°F (8.3 microstrain/10 K) and exhibits a zero drift as high as 40 microstrain over a period of 50 days. For cross-sectional areas equal to or less than 50 in.2 (323 cm2), the Telemac SC-11 gage shows instrument error less than 10 percent of the strut load, provided that the gage reading has been corrected for temperature and zero drift. For similar conditions, the Telemac SC-8 and Geonor gages show an instrument error of approximately 5 percent of the strut load. The design of strut load instrumentation is detailed under five headings which provide recommendations concerning choice of gages, installation, positioning, remote observations, and supplemental measurement techniques. Strut load measurements from a 60-ft (18.2-m) deep braced cut are analyzed with respect to the special characteristics of the Telemac SC-11 gages. The magnitude and distribution of brace loads in response to the removal of bottom struts from the excavation are summarized.