ABSTRACT Tomato (Solanum lycopersicum L.) is one of the most economically important vegetables in the world. Mexico is considered as its center of domestication and there is a large genetic diversity. Grafting in tomato has grown for various purposes including the increase of yield. An alternative use of native tomato genotypes is as rootstocks for grafting improved tomato. The objective of this work was to evaluate native accessions of tomato as rootstocks to identify outstanding genotypes for their potential to be used as rootstocks in tomato production. An experiment was conducted for two cropping cycles (2014 and 2015) in greenhouse and hydroponic conditions, in a randomized complete block design (RCBD) with three replicates and 10 plants per experimental unit. Treatments were formed by a combination of nine native tomatoes and two commercial rootstocks with two hybrids used as scions. Twelve traits were recorded and most of the treatments were significantly different (P < 0.05) from each other for these traits. The accessions LOR-22, LOR-77, LOR-81, LOR-84, LOR-95 and LOR-100 with the hybrid &#8216;El Cid&#8217;, and LOR-81, LOR-84 and LOR-100 with the hybrid &#8216;Sun 7705&#8217;, increased significantly yield by 19% and 22%, respectively, compared to ungrafted control. Moreover, characteristics related to fruit quality were preserved with grafting. The best combination scion/rootstock (&#8216;Cid&#8217;/100) yielded 30% higher than hybrid &#8216;El Cid&#8217; without grafting and 16% higher than the commercial rootstock &#8216;Multifort&#8217;. This allowed identifying genotypes of Mexican native tomatoes with great potential to be used as rootstocks or as source of germplasm for rootstock development.