Short-term deployments of a low-cost triaxial velocimeter were conducted to assess its suitability for ambient vibration testing and operational modal analysis of a large concrete gravity dam. Prior to field deployment, the velocimeter was collocated with industry-standard instruments in a controlled setting to facilitate verification of the recordings. The velocity time series were compared, and the low-cost instrument recordings exhibit significant differences compared with those recorded by the industry-standard instrument. Spectra were also compared in the frequency domain and the low-cost velocimeter was found to accurately record the dominant frequency content; however, these spectra are limited to frequencies below 50 Hz due to the low sampling rate of the velocimeter. The velocimeter was then deployed as a standalone instrument at a large concrete gravity dam for approximately 20 minutes of data acquisition at two locations: (i) a public pathway above the dam crest and (ii) at a free field approximately 100 m downstream of the dam. Estimates of resonant frequencies were made based on observed spectral shapes; however, confidence in these estimates was compromised by significant low-frequency drift, poor time synchronization between channels, and the possible presence of machinery excitation. Although the low-cost velocimeter may be suitable for certain applications, several important limitations are identified, which include: (i) apparent lack of reliable time synchronization, (ii) limited user interface which increases the likelihood of user error, (iii) reliance on an external power source which complicates field deployment, (iv) low sampling rate, and (v) greater susceptibility to instrument malfunction.