The recovery of microbial communities removed from hard substrates was studied in the laboratory and on microepiphyton in the White Sea, in St. Petersburg, and in Karelija. The removal of these communities from hard substrates destroyed their spatial structure. Recovery was allowed to proceed from the suspension on a horizontal substrate in stagnant water at low temperature. The sequence of colonization in the laboratory was the same as that in the field. Laboratory recovery was complete in 12-24 h and did not depend on the age of the natural community. From the comparison between laboratory and field data on such mechanical disturbance, a pattern of recovery, termed ‘self-assembly', is proposed.