Scientific workflows are groups of scientific application tasks organized in oriented graphs. These scientific workflows are characterized by a large number of tasks requiring sufficient resources for their execution. Tasks with all available input data can be computed simultaneously. Cloud computing is an appropriate environment for the implementation of scientific workflows. Although the cloud computing environment has unlimited resources and can run some scientific workflow tasks simultaneously, scheduling scientific workflow tasks using pay-as-you-go cloud computing resources is an NP-complete problem. This difficulty is due to constraints on the part of the cloud resource provider and the part of the user (customer). The algorithm tries to find efficient schedules that take into account several requirements of client service (QoS) such as deadlines, budgets, and resource providers’ profits, i.e., the minimization of energy consumption and many others. There have been several papers recommending effective solutions to workflow schedule problems. This paper reviews existing and more recent papers on the plan of scientific workflows in pay-as-you-go IaaS cloud computing environments, focusing on future directions for algorithms that can improve the optimal solution.