The ever increasing public demand and the setting-up of national and international legislation on safety assessment of potentially dangerous plant require that a correspondingly increased effort be devoted by regulatory bodies and industrial organizations to reliability data in order to produce safety analyses. Reliability data are also needed to assess availability of plant and services and to improve quality of production processes, in particular, to meet the needs of plant operators and/or designers regarding maintenance planning and production availability. The need for an educational effort in the field of data acquisition and processing has been stressed within the framework of EuReDatA, an association of organizations operating reliability data banks. This association aims to promote data exchange and pooling of data between organizations and to encourage the adoption of compatible standards and basic definitions for a consistent exchange of reliability data. Such basic definitions are considered to be essential in order to improve quality. To cover issues directly linked to the above areas, space is devoted to the definition of failure events, comon cause and human error data, feedback of operational and disturbance data, event data analysis, life-time distributions, cumulative distribution functions, density functions, Bayesian inference methods, mutivariate analysis, fuzzy sets and possibility theory.