A continuous observation during construction always has been an important element of geotechnical engineering for minimizing risks. The reasons are the uncertainties in the geological and geotechnical models, as well as the limited capability to sufficiently accurately describe the complex ground behaviour. In the early days design of geotechnical structures thus was mainly based on experience. As usual, experience was used to develop mathematical solutions to the problem. To account for the inherent scatter of parameters and uncertainties in the models, the assumptions made during design had and have to be verified or falsified during construction by observations. For safe and economical construction, the design has to be adjusted during construction to the actual conditions. The term “observational method” was first formally introduced by Peck and Terzaghi. The original ideas behind the method have not always been understood or followed. The paper, after a brief historical review of the observational method shows the current status of the method is critically reviewed, and further developments discussed.