Purpose of data capture: Records in this dataset were originally collected for a variety of purposes including: identification of areas of marine natural heritage importance, to support protected site designation or to support site condition monitoring. This dataset is a collation of species records from these surveys and has been produced to mobilise these records to a broad spectrum of users and in so doing it contributes to Scottish Natural Heritage?s ongoing commitment to the National Biodiversity Network. Methods of data capture: Records in the dataset were collected as a result of commissioned surveys which used a range of methodologies often closely aligned with those of the Marine Nature Conservation Review (MNCR) as described in the MNCR Rational and Methods report (Hiscock 1996). Surveys were typically undertaken using one or more of the following techniques, remotely operated video (ROV), drop-down or towed video, grab sampling, core sampling, clump sampling, diver video, MNCR phase 2 diver recording, and intertidal transects and quadrats. Most commonly semi-quantitative recording techniques have been applied using the SACFOR abundance scale, but species presence, and count data have also been used. Typically although not presented in this dataset the surveys would also have described the variety of habitats present (detailing substrata, depths, wave exposures, current regimes, salinity and so on). Verification and validation of records has typically focused on a quality assurance protocol for taxon determination and data entry. Geographical coverage: The records in this dataset relate to the Scottish Marine Area, but have a predominant near-shore focus in their distribution. Temporal coverage: The records present in this dataset relate to field surveys 1993 up to and including 2012. Data quality: The majority of records have been subjected to quality assurance procedures which includes an independent oversight of a sub-set of records from each survey and specialist taxonomic input. In addition some of the data has been subject to NMBAQC protocols. Additionally, SNH project managers? have an oversight on the survey deliverables. Overall confidence in the dataset is high although it is conceivable that errors exist in this dataset. SNH view mobilisation of these records to the NBN Gateway as a further mechanism for assuring data quality. If users of the Gateway identify potential errors in the dataset then these should be flagged with SNH who will consider all feedback.