In this article we present some of the more important findings from research aimed at the validation and normalisation of the Family Assessment Questionnaire (Dyadic Questionnaire, Family Questionnaire, Self-Estimation Questionnaire), which is the Polish adaptation of Manfred Cierpka's and Gabriele Frevert's "Familienboegen". In the presented study 1511 individuals from 557 families took part. Of these 658 individuals (including 162 children) came from 248 families which had no clinical health or adjustment problems, while 853 individuals (including 305 children) came from 309 families with problems due to schizophrenia, anorexia nervosa, coronary heart disease or family crisis. A new method of analysing raw results from the questionnaire scales was formulated and verified. This method was based on the criterium of sufficient satisfaction regarding the aspect of family life measured by the scale. The original scales of all the questionnaires (Task Completion, Role Performance, Communication, Emotionality, Affective Involvement, Control, and Values and norms) were characterised by average or low reliability; the general scales were characterised by high reliability (the Dyadic Questionnaire with very high reliability). As a result of factor analysis new scales were created with acceptable or high reliability. These were the scale of Positive statements, the scale of Negative statements in the dyadic and family questionnaires, and the scales of kindness, care and resentment in the Self-Estimation Questionnaire. The general scales and the factor scales generally significantly differentiated between the sub-samples selected due to the type of family problems present and relationship. On the other hand the 7 original scales did so only sporadically or weakly. Better results were obtained by families without health or adjustment problems while the worst results were obtained by families in crisis.