Promjene koje u komunikaciji među ljudima, a time i u obliku pisanih dokumenata, izazivaju informatičke tehnologije, čini aktualnim pitanje o sadržaju djelatnosti arhivske struke. Osnova te djelatnosti je upravljanje spisima tijekom čitava njihova životnog ciklusa te je stoga nužno jasno razumjeti što je spis, odnosno dokument i što ga razlikuje od drugih oblika informacija. Arhivistička literatura poznaje dva različita tipa definicije dokumenta: pravni i teorijski. U članku se opisuju dva pristupa definiranju elektroničkih dokumenata u kontekstu istraživačkih projekata u posljednje vrijeme: jedan koji definiciju izvodi iz poslovnih transakcija, i drugi, s uporištem u diplomatici, koji polazi od tradicionalne definicije arhivskog dokumenta, pri čemu se specifičnost elektroničkog dokumenta vidi u fizičkoj odvojenosti teksta i kontekstualnih podataka. Iz definicije i svojstava dokumenta izvode se i funkcionalni zahtjevi sustava za upravljanje spisima i dokumentima., More and more people communicate by electronic mail, in stead of calling, and by sending and receiving c-mail messages they create potential records. It looks like that again orality is loosing its position in social life. Every peace of recorded information is potentially a record. Today in some countries, in the near future in other countries, archives will be surrounded by the electronic records, and overwhelmed by a continuously increasing of records production. These forthcoming changes, these challenges, make us put ourselves the eternal question; What business are we in. To give at least one answer: We arc in the business of keeping records. That is our core business, and we might be in some other related business, such as in the information business, but it is all about records. Not about current records, or semi-current records or non-current records - no, just records. Since we are in the business of keeping records, we need a clear understanding of what a record is, and what distinguishes a record from other pieces of information. In the current archival literature on this subject, we sec two different perspectives for defining electronic records: a legal one, often chosen by archives as a basis for interventions in public administration, and an academic one, chosen by universities, a basis for fundamental archival research. Looking more closely at the academic perspective, we may discover two different approaches. The University of Pittsburgh, for example takes the business transactions of organisations as a starting point. A business transaction creates, and uses records. The records arc the evidence of the transactions that created and used them. Contrary to Pittsburgh the UBC project provides with clear definitions of electronic records, and in a structured, in diplomatics founded way. The starting point for the definition of an electronic record is the definition of a document, than of an archival document. An archival document is not a document with archival value, in the sense that it is worth to be preserved permanently, but a document which has re-cordness, played a role in a business process, and may serve as evidence of that process, An electronic document is a document created and communicated in a digital format and by computer technology. The advantage of the UBC definition is that it links up with traditional thinking, although that might be a risk as well. At least for the short term the use of the concept of the document as a metaphor has advantages for understanding. Whereas Pittsburgh appears to be more dynamic and open for any kind of emerging technology, it boars the risk of being one step too far. Defining a record as a document, and an electronic record as an electronic do¬cument implies that one characteristic of both an electronic and a traditional record is that it must be complete, intelligible in itself. One characteristic is according to the UBC project most essential for any kind of record: its interrelationships. A record can only be fully understood in conjunction with other records, for instance being part of a scries or a case file. Archivists are first of all responsible for keeping records, therefore a major question is how electronic records must be kept. The assumption that electronic records must be kept electronically, contrary to printing them out on paper, find a strong support by many archival thinkers, because of the recognition that a record created, communicated and used electronically, finds its authentic form in its origi¬nal electronic format. For this very reason it is needed for each legal, political and societal system to define functional requirements forrccordkecping. Starting with the identification of the required quality of the records any recordkeeping system must preserve these required quality. It must be able to keep the records complete, reliable, authentic. It must be able to protect the records against change, illegal access, unwanted deletion. It must preserve information about the context in which the records have been created, information about who created the record, when it was communicated, when it was read, and in the course of what business process. These requirements derive from archival theory, from diplomatics, from the legal systems, from political demands, societal behaviour, financial regulations. One other functional requirement about which the opinions seem to agree, is that the recordkeeping system has to be or to become pro-active, not waiting for records eventually to enter into the system, but avoiding any risk of having records be lost, or violated. Recordkeeping systems, unlike information systems, provide time bound, non-manipulable, and highly redundant information. A recordkeeping system is not apiece of software, an application. It is more than that, it is the whole of procedures, rules, knowledge, hardware, software, tools, methodologies, and people, including the records themselves of an organisations, preserving them and making them available for use by providing access for those who have the rights to access them.