Field sampling carried out in the Rfa de Vigo (NW Spain) from 1991 to 1993 showed that Mytdus galloprovincialis settle directly from the plankton onto substrates exposed to various environmental conditions: exposed rocky shore; protected rocky shore; exposed, raft mussel-culture area; and protected, raft mussel-culture area. For direct settlement, competent larvae may use a wide variety of substrates: filamentous nylon ropes; the byssus and intricately arranged material in the bottom of mussel beds; filamentous, thallus and membranous algae; and rugosities on adult mussel shells. The peak of settlement occurred from spring to early autumn and differences in the settlement abundance among localities were influenced by currents. After this peak, the settlement of larger post-larvae continued, associated with the increase in storms during autumn that detach them from their original substrates. This dispersion phase allows for the possibility of colonising, or recruiting on other areas, even during the post-spawning season when the presence of small post-larvae is at its minlmum. Laboratory experiments carried out with post-larval stages from 0.250 to 2.000 mm showed that under static water conditions they crawl and form clumps, but do not search for a specific substrate. If they are not disturbed, they may remain in their original place of settlement. Conversely, under moving water conditions they attach to a wide variety of substrates, particularly to byssal filaments and thalli of red algae Ceramium rubrum. The contact and attachment to substrates is carried out with a long mucous thread that also aids in forming clumps. The use of this mucus to settle results in a 'preference' for natural filamentous substrates but also in settlement on rugous hard surfaces. An alternative hypothesis to the primary and secondary settlement pattern previously described in the literature for Mytilus edulis is suggested.