This data publication includes data on catches reported in the harbours of Santa Catarina state, southern Brazil, between the years 2000 and 2019. Historically, these harbours have concentrated a significant part of the demersal fishing fleet that operates widely on the Brazilian Meridional Margin (BMM), from 21°S to the southern border of Brazilian EEZ (34°S) and from the coastal areas down to 500 m depths. Landed catches were monitored by a research team based on the University of 'Vale do Itajaí' (UNIVALI) along a sequence of scientific projects and contracts developed to meet governmental demands for oceanic and deep fisheries development and management, and in support of the licensing processes of the offshore oil and gas exploration activities. This catch composition time series shows an annual compilation of total catches discriminated by species caught by trawlers (single trawlers, pair or twin trawlers and double trawlers) and gill net industrial fisheries. These species represent altogether over 81% of total reported catch in the period. The remainder amount of landings was reported as groups of species (e.g. sharks) or categories indiscriminated by species and were excluded from the database.The 'Mean Temperature of the Catch (MTC)' was estimated for each year of the time series, as proposed by Cheung et al. (2013). This variable is expressed in degrees Celsius. 'Sea Bottom Temperature (SBT)' was derived from estimates provided by the high-resolution ocean general circulation INALT20 model (Schwarzkopf et al., 2019) for the study period (2000-2019) and was calculated by averaging the temperatures over 0.25° x 0.25° grid cells of the BMM and a water column up to 50 m above the seafloor. This variable is expressed is degrees Celsius. 'Annual volume transports of the Brazil-Malvinas confluence (BCt)' was extracted and compiled from Artana et al. (2019). This index was estimated for the period between 2000 and 2017. More detailed information about this index can be accessed in Artana et al. (2019). This variable expressed in Sverdrups (Sv). 'Simpson diversity index (Dm)' express the diversity of demersal fishing métiers (i.e. combination of target species, gear, and time of the year) operating each year of the time series, as determined by a process of classification of individual fishing trips. Catches of each métier operating during each year of time series summed and used to calculate Dm using the Simpson species diversity index formulation. The temperature preference/affinities data was obtained from global compilations made available by Cheung et al. (2013) and in FishBase (Froese and Pauly, 2022). In both compilations, thermal preferences derive from considerations about the species distribution ranges and sea surface temperature maps (e.g. Cheung et al., 2013 supplementary material). Each species considered here was assigned global species identifications or references like FAO ISSCAAP, FAO TAXOCODE, FAO CODE ASFIS 3A, AphiaID, and ITIS/TSN.