In the last two decades, all earthquake hazard studies in Spain have been done using the attenuation deduced from the isoseismal maps of the most important earthquakes. However, it is well known that the preferred parameter to measure the size of any shock is magnitude and, consequently, the knowledge of acceleration as a function of magnitude and distance is basic to any new hazard determination. The installation of a strong-motion network in Spain, with more than a hundred digital instruments set up in the last 15 years, made possible the recording of more than 447 shocks in the moment magnitude range of 1:3 ≤ M ≤ 6:3. However, only 149 shocks were processed corresponding to magnitudes greater than 3 and a distance of less than 100 km. Values of the maximum horizontal acceleration recorded on firm soil permitted us to make a regression analysis, which has been compared with similar previously published relationships. Additionally, the southwestern part of Spain is also affected by a marine source area of different very low attenuation, but there are not enough recordings to attempt a regression analysis. In this case, the only strong-motion data gathered for the area was checked against the two intensity attenuation relationships available after conversion of maximum intensity to moment magnitude, and intensity to acceleration, showing a reasonable behavior of the newest intensity attenuation law gathered for the area.