These data comprise a grain character dataset (grain size, sorting and grain shape) from the topset, foreset, and bottomset deposits of four successive Miocene intrashelf clinothem sequences (m5.7, m5.4, m5.45 and m5.3). These clinothems have been mapped and described by various authors (e.g. Monteverde et al., 2008; Mountain et al., 2010; Miller et al., 2013), and were continuously cored and logged during IODP (International Ocean Discovery Program) Expedition 313 (Offshore New Jersey, USA). In total, 878 sediment samples were collected from the working half of three cores recovered during IODP Expedition 313, offshore New Jersey. The three cores, kept in cold storage at the University of Bremen, are from Sites M27, M28 and M29. The stratigraphic horizons targeted during this investigation were exclusively Miocene in age, corresponding to depths of 225 - 365 mcd (metres composite depth), 312 - 600 mcd, and 600 - 730 mcd in cores M27, M28 and M29 respectively. Collectively, a total of 560 m of core has been sampled. With reference to the seismic clinothem model presented in Miller et al. (2013), these stratigraphic depths correspond to the interval between major seismic sequence boundaries m5.7 - m5.2. Where no prominent grain size change was recorded in either the cumulative lithology presented in Miller et al. (2013) or core descriptions (Mountain et al., 2010), the strategy for sample collection was to remove 15 x 15 x 15 mm sediment slices, subsampled at ~ 500 mm intervals down-core. The sampling strategy was amended to target stratigraphic depths where grain size change was most prominent. At these intervals, highlighted by the broad patterns of down-core lithological and grain size change (Mountain et al., 2010; Browning et al., 2013; Miller et al., 2013), sampling density was increased to ~ 300 mm intervals. During the sampling process there was some deviation from this sampling configuration in order to avoid 1) horizons of cementation, (2) biscuiting disturbance, 3) key stratigraphic surfaces and 4) heavily sampled intervals. Due to the pervasive presence of biogenic material (including calcareous skeletal remains, shell fragments, and organic matter) sample pre-treatment was undertaken prior to grain character measurements, in order to remove these components. Sample pre-treatment comprised the careful manual disaggregation of the semi-lithified samples using an agate mortar and pestle (e.g., Sahu, 1964; Wilson and Pittman, 1977; Nelson, 1983; Frey and Payne, 1996; Ando et al., 2014). Hydrochloric acid (10% weight to volume) and hydrogen peroxide (30% weight to volume) were added to all samples, to ensure the removal of any calcareous and non-calcareous organic components, respectively (e.g., Battarbee, 1986; Battarbee et al., 2001; Gray et al., 2009) Grain character is defined as the grain size, sorting and grain shape (sphericity and roundness) of a sample. Grain character analysis was completed using a CamsizerXT (Retsch Technology), which is an optically based dynamic image analyser. The CamsizerXT is capable of measuring the grain-size range 1 µm - 8 mm (clay - gravel), with an accuracy of ± 1%. Grain-size fractions < 1 µm are lost during the process of analysis. The statistical analysis of all CamsizerXT results was completed using GRADISTAT computer software (Blott and Pye, 2001). The GRADISTAT software enables the rapid analysis of grain size statistics from multiple sediment samples and produces numerical, geometrically and logarithmically calculated values of the mean, mode, and sorting (more information on Page Two of data spreadsheet). Grain shape data were analysed using Microsoft Excel software. The data are presented in spreadsheet format. Each sample is given a 'Site' this refers to Sites M27, M28 and M29; for the site locations the user is asked to refer to the seismic clinothem model presented in Mountain et al. (2010) and Miller et al. (2013). Each sample is also given a core number and core section (e.g., 80-1) and a sample depth (given in meters composite depth; e.g., 225.55 mcd); the user is asked to refer to Mountain et al. (2010) and Browning et al. (2013) to see the sampled core numbers, sections and depths alongside sedimentary logs, completed by the Onshore Scientific Party of IODP Expedition 313. Each sample is given: i) a textural group and a sediment name; ii) an arithmetic, geometric and logarithmic value of the grain-size mean, standard deviation, skewness and kurtosis; iii) an arithmetic, geometric and logarithmic value of the sorting and iv) the percentage of each grain-size fraction present within a sample. On Page Two of the spreadsheet (entitled GRADISTAT Information); Tables One and Two outline the mathematical parameters and grain-size scale used by the GRADISTAT software (Blott and Pye, 2001) to calculate the grain character data. For more information the user is asked to refer to Blott and Pye (2001). A synthesis of the grain character results and interpretations have been published in two papers; Cosgrove et al. (2018) and Cosgrove et al. (2019). These papers document patterns of sediment dispersal and variations in grain size, sorting and grain shape, at a basin-scale (i.e. across successive clinothems) and within individual clinothem sequences (i.e. at an intra-clinothem scale). These data can be used to condition and validate process-based numerical forward models and have widespread applications in prediction of reservoir quality in both frontier and mature hydrocarbon basins.