In-field data were collected in Costa Rica between 2018–2021 on newly planted grafted and non-grafted coffee plants grown under artificial shade nets and across an elevation gradient (1050, 1250 and 1450 m.a.s.l). The coffee plants consisted of Coffea arabica F1 hybrid plants (‘H3 i.e. Caturra cv. X Ethiopian 531’), which were derived from a somatic embryogenesis clonal propagation process, an American C. arabica pure line (‘Villa Sarchi’) and C. canephora ‘Nemaya’ (the latter two both being produced by seed). Data from eight different coffee types (including these three genotypes) and different grafting combinations (including reverse and auto-grafting) were collected. Data concerned plant traits such as grafting compatibility (plant collar diameters above and below graft union), agronomic characteristics (aerial and root traits), leaf ecophysiology (leaf gas-exchange and chlorophyll fluorescence), yield and quality attributes (bean size, peaberry percentage, WB100 and SCA note). Climate data were also included for comparison on the farm plots along the elevation gradient. Linear mixed models were used to test for effects of elevation (test sites), coffee types (grafted or non-grafted combinations) and interaction between coffee types and elevations. Least square mean estimates were calculated for significant fixed effects and Tukey tests applied for pairwise tests. A tangential hyperbola curve was used to analyse leaf gas-exchange data. These datasets and R scripts can be re-used as a guide for future analyses concerning coffee agronomy or eco-physiological interactions for other plant species. Other potential re-uses could be meta-analyses aimed at comparing coffee yield, quality, or other agronomic traits across different environmental conditions (such as under shade of an agroforestry system or across different elevation sites).