The SH waves from 21 underground nuclear explosions at Pahute Mesa (NTS) were used to investigate tectonic release. The equivalent double-couple representation of the tectonic release, which was constrained by waveform modeling and the polarity of sP, is very similar for all the explosions. The average orientation is a right-lateral, strike-slip fault trending N15°W. Seismic moments were determined on the basis of comparisons with two western United States strike-slip earthquakes. BENHAM has the largest tectonic release moment (5.6 × 1024 dyne-cm) and STILTON (0.1 × 1024 dyne-cm) the smallest of the events studied. In general, the seismic moments increase with the size of the explosion, but the location of the explosion relative to previous explosions can strongly affect the tectonic release. The Pahute Mesa events can be separated into two populations: (1) events which are well separated (>4 km) from previous explosions, and (2) those events which are close (<4 km) to previous explosions. Those events which are close to previous explosions show a marked decrease in tectonic release. A least-squares fit of seismic moment to event size (as determined from the world wide, average ab amplitude) shows that the two populations are approximately parallel, but offset. The fact that the trends remain separated even at small yields (e.g., PIPKIN and SCOTCH) suggests that for Pahute Mesa there is not a threshold for tectonic release. Since spatial position gives the best separation of high and low tectonic release events, a volume model is favored for the source mechanism. The preferred model is motion on a system of faults and joints.