This lecture starts with a survey of the present state of knowledge on the topology of potential hypersurfaces of proteins, on transitions in conformational space, and on harmonic and quasi‐harmonic methods of analysis of sub‐state properties. The concepts of deterministic vs ergodic behaviour of the paths actually followed in conformational space are discussed. The possible connection between selectivity (or ‘‘accuracy’’) in enzyme catalysis and the time structure of conformational transitions is shown. In a second part, two different methods used by the authors for theoretical computation of conformational paths linking known crystallographic structures in citrate synthase are commented: an implementation of Elber and Karplus’ chain algorithm, and a new method which we named directed dynamics. In connection with the ‘‘multiple minima problem’’, the occurrence of lower and upper bounds in the temperatures used in molecular dynamics searches of conformational minima is discussed with emphasis on their dependence with respect to the effective number of degrees of freedom in the conformational space investigated. Finally, a hint on the total transition time is given.