By ignoring the effects of nonlinear propagation, current exposimetry protocols may yield significant underestimates of the acoustic pressure in situ. This problem can be avoided simply by (1) extrapolating pressures linearly from low amplitude measurements in water and (2) linearly derating these values to obtain estimates of fields in situ. The mechanical index was designed to provide an indication of temporal peak acoustic fields for use in prediction of nonthermal biological effects in tissues. At low outputs, the mechanical index, together with the frequency, gives the peak negative pressure near the focus of the field. As currently formulated, however, the pressure used in the mechanical index may be far from the focus at high output levels. Recommendations of the World Federation of Ultrasound in Medicine and Biology avoid the underestimate associated with nonlinear propagation as well as other problems with the mechanical index and may be preferable in dealing with non-thermal bioeffects. The thermal indices that are implemented currently in the Output Display Standard (American Institute of Ultrasound in Medicine/National Electrical Manufacturers' Association) are affected less seriously by nonlinear propagation.