As part of the EUROCarbDB project (www.eurocarbdb.org) we have carefully analyzed the encoding capabilities of all existing carbohydrate sequence formats and the content of publically available structure databases. We have found that none of the existing structural encoding schemata are capable of coping with the full complexity to be expected for experimentally derived structural carbohydrate sequence data across all taxonomic sources. This gap motivated us to define an encoding scheme for complex carbohydrates, named GlycoCT, to overcome the current limitations. This new format is based on a connection table approach, instead of a linear encoding scheme, to describe the carbohydrate sequences, with a controlled vocabulary to name monosaccharides, adopting IUPAC rules to generate a consistent, machine-readable nomenclature. The format uses a block concept to describe frequently occurring special features of carbohydrate sequences like repeating units. It exists in two variants, a condensed form and a more verbose XML syntax. Sorting rules assure the uniqueness of the condensed form, thus making it suitable as a direct primary key for database applications, which rely on unique identifiers. GlycoCT encompasses the capabilities of the heterogeneous landscape of digital encoding schemata in glycomics and is thus a step forward on the way to a unified and broadly accepted sequence format in glycobioinformatics.