In the paper a construction scheduling problem – namely flowshop – with minimizing the sum of penalties for exceeding the deadline of building structures completion is considered. The problem is illustrated by the investment task concerning the implementation of twelve apartment buildings forming a part of a new housing estate. Uncertain parameters of the system are represented either by fuzzy numbers or random variables, whereas random variables have normal or the Erlang distribution. Since even the deterministic version of the problem is strongly NP-hard, the approximate algorithm based on the tabu search method was used to its solution. The performed computational experiments showed large solution resistance against any potential interference of parameters of the problem.