Results are presented of field research on water distribution in the command area, covering 18,200 ha, of a secondary irrigation canal in the Tungabhadra Left Bank Scheme, Karnataka Stare, India. The official objective of the Scheme and the resulting implications for the water distribution are discussed first. An explanation of the planning and operation of the water distribution follows. The results are based on analyses of the water flows taken from the D36 secondary canal and distributed along the canal to the "pipe outlets" (inlet structures to the tertiary units), and of the canal section rotation practised along the canal. The analyses concentrate on three dimensions of the water supply: - The design flows, according to the official Scheme objectives and criteria; - The targets, as set by the system operators before every season; - The actual distribution procedures and flows, as observed during the operation. The analyses, supported by flow measurement data, illustrate that the water distribution is not based on consistent and clear criteria and procedures, but that it is the outcome of varying compromises, decided upon pragmatically by the field staff, to bridge the gap between the farmers' demands and the upstrearn constraints to water availability. This paper explains the widespread phenomenon of head reaches taking too much water, leaving little or nothing for the tail end of the canal.