The aims of the present empirical study, based on questionnaires filled in by first-time pregnant women, have been to build a complex explanatory model for the period of pregnancy, childbirth and very early infant-caregiver relationship. The main question was whether individual- (e.g. trait anxiety) and psycho-social factors (e.g. own childhood, female sex-roles, support from the partner, desire for the baby) were predictive of interactive characteristics of early postnatal relationship. The present analyses using the method of path-analysis have searched for direct links and indirect paths through the attitude and anxiety towards pregnancy or experiences of childbearing. Data were collected longitudinally by questionnaires from a controlled sample of 78 first-time pregnant women, so the results cannot be generalised for the population. The 13 variables of the model were obtained by principal component-analysis of our own questionnaires and from scales of other standardised and publihed questionnaires (STAI, PAS, MABS). In some cases, data reduction was done to create groups of variables. The statistical model was constructed on the basis of the relationships between the variables of the two study periods, the 26-36 weeks of pregnancy and the first 2 weeks of very early postnatal relationship. Variables at the first level of the model were trait anxiety and psycho-social factors. The second level included anxiety and attitude towards the pregnancy, and at the third level there were the variables describing childbearing experiences and very early dyadic contact after the childbirth. The last, fourth level of the model included the factors of irritability and sociality of the baby, self-confidence of the caregiver, and satisfaction with the changed life-style after childbirth. Without detailing the paths it can be concluded that the parameters of the model (good fit-indexes, significant paths, the explained variance scores of the variables) support the modelled prediction that the early self-confidence of the caregiver can be predicted not only directly by the perceived irritability of the child (as other researches have already shown), but also indirectly by prenatal factors mediated by the attitude towards pregnancy. Thus, prenatal factors may have protective effects counterbalancing risks carried by negative child characteristics.