Previous in vitro studies of acoustic coupling in shock wave lithotripsy (SWL) have shown that air pockets trapped at the surface of the treatment head significantly reduce transmission of shock wave (SW) energy to the focal zone of the lithotripter, reducing the effectiveness of stone breakage. Since there are no reliable means to monitor the quality of coupling during SWL, we looked for a practical protocol to improve how coupling is achieved. In vitro studies were performed using a Dornier DoLi-50 lithotripter. LithoClear gel was used to couple the treatment head to the acoustic window of a clear acrylic test tank. Numerous methods of applying gel were tested including common sense variations of routine protocols typically used with patients. For each method the coverage of air pockets (% defects) was determined using digital imaging. Different coupling regimes were tested for effect on the breakage of gypsum model stones. The quality of acoustic coupling was affected by how the gel was handled--how it was dispensed and applied, and whether the gel was applied only to the treatment head or to both the lithotripter water cushion and the test tank (surrogate patient). Dispensing gel from a squeeze bottle for application by hand created significantly more defects than when a large volume (approximately 250 ml) of gel from the stock jug was applied as a mound to just the treatment head (26.5+/-2.7 vs. 1.2+/-0.5% defects, P<0.001). The efficiency of stone breakage was better when gel was applied from the stock jug compared to application by hand (P<0.006). Poor coupling was substantially improved by using the inflation feature of the water cushion to collapse air pockets, but this strategy was not a substitute for establishing good coupling at the outset. The quality of coupling in shock wave lithotripsy can be improved by minimizing the handling of the coupling medium. Hand application of coupling gel is clearly not the best way to prepare for lithotripsy. Better results can be obtained by delivering lithotripsy gel as a bolus to the treatment head alone, and allowing it to spread upon contact between the treatment head and the skin. These in vitro tests also suggest that the inflation feature of the lithotripter may be useful in reducing defects in coupling.