Direct measurement of seismic attenuation (1/Q) and velocity dispersion in the laboratory point to several independent loss mechanisms. A forced-deformation technique allows measurement at seismic frequencies and amplitudes. Fluid motion is the primary mechanism in porous, permeable clastics. In shales, bulk-fluid motion is inhibited, and clay-particle interaction with bound water might dominate. Another mechanism usually not considered is losses in the fluid phase alone. Heavy, viscous fluids show bulk losses independent of a rock matrix. All these loss mechanisms are frequency dependent, so observations of 1/Q made at seismic frequencies usually will not agree with ultrasonic data. Attenuation and velocity dispersion are coupled processes, and validity of the frequency-dependent 1/Q values can be checked by comparison with observed velocity dispersion over a broad frequency range.