Many heavily fished coral reefs have a high abundance of sea urchins that may suppress the recovery of fish and coral populations once fishing effort is reduced or eliminated. Restoration of these reefs may be accelerated by intervening and reducing sea urchin populations—particularly those of long‐lived species. We studied three Kenyan coral‐reef lagoons to determine the influence of reductions in sea urchin populations on coral, algae, and fish populations. Populations were monitored seven times over a 1‐year period in 50 × 50 m unmanipulated control plots and experimental sea urchin reduction plots where sea urchin populations were reduced by about 85%. Census of the most‐abundant fish families found the greatest positive population responses to sea urchin reduction in plots protected from fishing. The wet weights of fish nearly tripled, population density increased by 65%, and species richness increased by 30% compared with adjacent control plots. Parrotfish, wrasse, scavenger, and snapper families showed the greatest population and wet‐weight increases. In the two fished reefs, fish populations also increased but to a lesser degree than in the unfished reef—particularly when comparing the wet‐weight estimates of fish. Small‐bodied species of the dam‐selfish and wrasse families and juvenile parrotfish exhibited the largest population increases in these fished reefs. In fished reefs algae and seagrass cover exhibited the greatest increases following sea urchin reduction. Seagrass colonized bare sand, whereas fleshy brown algae colonized hard substrate (dead coral). The tall canopy‐forming alga Sargassum latifolium became the dominant alga in the fished reefs, whereas the more prostrate genus Padina dominated in the unfished reef. Fleshy algae cover in the unfished reef was about half that of the two fished reefs—attributable to the greater abundance of parrotfish in the unfished reef. This suggests that greater herbivory in the unfished reef kept the algae from reaching an algal forest climax typified by Sargassum dominance. In the fished reefs the high S. latifolium cover and the increased thickness of the algal turf reduced hard coral cover by around 30%. In the unfished reef coral cover was reduced by around 13%, but by the end of the experiment coral cover and genera richness were the same in both the control and the sea urchin reduction plots. We conclude that, primarily, fishing and, secondarily and indirectly, high sea urchin abundance are reducing fish numbers and diversity in Kenyan reefs and that sea urchin reduction has the potential to increase reef fisheries production and recovery from overfishing. Sea urchin reduction is not recommended, however, on moderately fished reefs due to the possible loss of coral cover and diversity. If reefs are either severely degraded or if fishing effort is reduced or eliminated, then sea urchin reduction is recommended for the restoration of fish numbers, feeding importance, species diversity, and fisheries production.