- This paper describes several measures that can be useful when analyzing features of search in genetic algorithms. Results obtained by different researchers are often hard to compare and it is often difficult to draw general conclusions from them. This means that non-experienced users of genetic algorithms have little help in constructing their own effective solutions for a particular problem. The article contains a suggestion about possible ways of investigating features of search in genetic algorithms. These features are defined together with descriptions how to measure them. © Springer-Verlag Berlin Heidelberg 1999.