I stole the MARC DB schema that I am using with Koha from somebody. I'd love to attribute, but I can't find it right now. It'll be attributed back in the koha-devel archives somewhere. I have done some work towards integrating this schema with the existing Koha database. My goal was to create an API that would abstract out the underlying database, so that changes made to a record could be made to either the old Koha style database or the new MARC database. My testing has shown the schema given below to be pretty poor at lookups. Some kind of separately maintained indexes would almost certainly be necessary. # Resource Table holds information about a particular resource Resource_Table Resource_ID | int(10) unsigned auto_increment primary key Date_Added | timestamp(14) Date_Modified | timestamp(14) Record_ID | int(10) unsigned # Bib_Table holds a record for each Tag in the MARC record. Links # back to Resource_Table with the Record_ID field Bib_Table Bib_ID | int(10) unsigned Record_ID | int(10) unsigned Date_Added | timestamp(14) Date_Modified | timestamp(14) Control_ID | int(10) unsigned Tag_0XX_ID | int(10) unsigned Tag_1XX_ID | int(10) unsigned Tag_2XX_ID | int(10) unsigned Tag_3XX_ID | int(10) unsigned Tag_4XX_ID | int(10) unsigned Tag_5XX_ID | int(10) unsigned Tag_6XX_ID | int(10) unsigned Tag_7XX_ID | int(10) unsigned Tag_8XX_ID | int(10) unsigned Tag_9XX_ID | int(10) unsigned # And for each of #XX, there are two tables, 6XX_Tag_Table and # 6XX_Subfield_Table. # # 6XX_Tag_Table holds a record for each 600s Tag in the MARC record. # Tag_ID links back to the Tag_6XX_ID field 6XX_Tag_Table Tag_Key | int(10) unsigned auto_increment primary key Tag_ID | int(10) unsigned Indicator1 | char(1) Indicator2 | char(1) Tag | char(3) Subfield_ID | int(10) unsigned Authority_ID | int(10) unsigned Link_Flag | enum('Y','N','B') Storage_ID | int(10) unsigned # 6XX_Subfield_Table holds a record for each subfield in each tag. # Subfield_ID links back to Subfield_ID in the 6XX_Tag_Table 6XX_Subfield_Table Subfield_Key | int(10) unsigned auto_increment primary key Subfield_ID | int(10) unsigned Subfield_Mark | char(1) Subfield_Value | char(255) Storage_ID | int(10) unsigned # The storage table is used for storing data that will not fit in a # standard 255 character database field. The Storage_ID field is used to # link to 6XX_Subfield_Table's Storage_ID, for example. Storage_Table Storage_Key | int(10) unsigned Storage_ID | int(10) unsigned Blob_ID | int(10) unsigned Text_ID | int(10) unsigned Med_Blob_ID | int(10) unsigned Med_Text_ID | int(10) unsigned Long_Blob_ID | int(10) unsigned Long_Text_ID | int(10) unsigned URI | char(255) Storage_Type | enum('B','MB','LB','U') The schema I stole had a slew of other tables dealing with Authorities and Holdings that I haven't documented here. Steve.