Procedures for impact assessment, including "beyond-BACI" (before-after control-impact) and proportional differences (ratios between impact and control treatments) were used to test population replenishment of marine invertebrates at a marine conservation area (MCA) and three fished (control) areas in the Solomon Islands of the southwestern tropical Pacific. Within shallow reef terrace habitat, the MCA caused abundance and size of the topshell Trochus niloticus to increase but did not affect holothurians (sea cucumbers) or the giant clam Tridacna maxima. Abundance of the nonexploited topshell Tectus pyramis was unchanged at the MCA but increased at the controls, possibly because of changes in abundance of T. niloticus. Within deep slope habitat, the MCA caused increased abundance of the sea cucumber Holothuria fuscogilva and prevented possible declines in abundances of Thelanota anax and all holothurians combined but had no effect on abundances of Holothuria atra or Holothuria fuscopunctata. Power analysis comparing the MCA with controls indicated that further, relatively modest increases in abundance or size of some species would have a good chance of being detected statistically. The beyond-BACI procedure holds promise for enabling rigorous evaluation of marine reserves as management tools at different spatial scales; the use of proportional differences is simpler but has limited management value.