Considering the development of computers and information technology, the present reliability assessment concept and format of standards for structural steel design seem to be outmoded. With designers using probabilistic reliability assessment concepts, Monte Carlo techniques and databases as tools, a reconstruction of the present standards can be expected. This paper addresses one possible alternative for qualitative changes in design standards and corresponding reengineering of the entire reliability assessment procedure. This includes introducing a probability of failure as a measure of structural reliability, developing a new generation of “hybrid” standards (basic and secondary standards, databases and software mutually interconnected using information nets), expressing variable input data as histograms and using simulation techniques to analyze the complex probabilistic interrelationships between load effects, material properties, and other quantities affecting reliability. The significance of the transition from a deterministic to probabilistic “way of thinking” is emphasized in terms of both education and design practice.