Almost ten years have passed since the idea of stellar associations as nonstable stellar systems was formulated. The complex of observational data obtained during this time indicates that stars contained in the associations are young objects of some million years of age. We would like to stress here that this concerns both the O and T associations. It is also known that those O associations which could be sufficiently investigated in this respect, contain, as a rule, T Tauri type stars and are consequently T associations as well. There are, on the other hand, T associations which do not contain hot giants. But apparently the mechanisms of stellar formations must be similar in O and T associations. This means that any theory of stellar origin for a given type of association must permit variations, which will provide an explanation of the origin of stars in associations of other type.
Two hypotheses on the origin of stellar associations have been thus far discussed. One of them, suggested by the author at the initial stage of the idea about associations, supposes that each association has originated as a result of an expansion from a body or a system, the volume of which was initially very small. The dimensions of the latter was in any case less than one parsec. According to this point of view, these initial bodies (protostars) have either not been observed up to the present, or have not yet been identified with any known object. This point of view does not give any indication about a concrete mechanism of stellar origin, postponing its explanation to the time, when the earliest stages of the expansion of the association may be studied in detail.