191 Introduction The sweet corn (Zea mays L. var. saccharata Sturt) is a vegetable intended solely for human consumption and can be marketed in natura, as green corn, or industrially processed in the form of canned food (Luz et al. 2014; Vendruscolo et al. 2018). The average productivity of grains of this culture is about 11 t ha -1 and approximately 36 thousand hectares are cultivated with sweet corn in Brazil (Luz et al. 2014; 2015). The interest in the culture also has grown due to its profitability and consequently becoming an important source of income to various producers (ABCSEM, 2014). All of these factors motivate the research by more productive sweet corn cultivars and more adequate genotypes to the requirements of the market. The main obstacle to the diffusion of sweet corn cultivation is related to the lack of cultivars adapted to tropical regions (Pereira Filho and Teixeira, 2016). The sweet corn cultivars must also achieve characteristics that meet the expectations of consumers, producers and the industrial sector (Terres et al. 2015); for such, it is imperative the use of tools for the selection of improved sweet corn cultivars. These selections are then based on the variables that meet the different expectations. However, selecting superior progenies is not an easy task because many of the characteristics of agronomic importance are, usually, of low heritability (Cruz 2013). The process of selecting superior genotypes can be more J. Crop Sci. Biotech. 2020 (March) 23 (2) : 191 ~ 196 DOI No. 10.1007/s12892-019-0334-0 RESEARCH ARTICLE Prediction of Genetic Gain in Sweet Corn using Selection Indexes Isadora Gonçalves da Silva 1 , Renata Castoldi 1 , Hamilton César de Oliveira Charlo 2 , Mateus de Souza Miranda 1 , Thaíssa Dias Cardoso Nunes 1 , Luciene Lacerda Costa 2 , Ernane Miranda Lemes 3* 1 Instituto de Ciências Agrárias (ICIAG), Universidade Federal de Uberlândia (UFU), Campus de Monte Carmelo, Zip code: 38.500-000. Monte Carmelo, Brazil 2 Instituto Federal de Educação, Ciência e Tecnologia do Triângulo Mineiro, Campus Uberaba, Zip code: 38.064-790. Uberaba, Brazil 3 ICIAG, UFU, Campus Glória, Zip code: 38.400-000. Uberlândia, Brazil Received: December 04, 2019 / Revised: December 10, 2019 / Accepted: December 31, 2019 Ⓒ Korean Society of Crop Science and Springer 2020 Abstract The cultivation of sweet corn is expanding in Brazil, but there are serious constraints about the availability of commercial cultivars. The selection of superior sweet corn genotypes can be performed based on selection indexes based on plant agronomic characteristics. Thus, the objective of this work was to compare selection indexes to select sweet corn genotypes aiming greater productions. The experiment was conducted at the Vegetable Experimental Station of the Federal University of Uberlândia, Campus Monte Carmelo. Eighteen sweet corn accessions (F3 generation) were evaluated. The selection indexes applied to the agronomic characteristics of sweet corn were: direct and indirect, the sum of ranks, desired gains and ideotype indexes. The characteristics evaluated presented significant differences among genotypes, except for stem diameter, prolificacy and grain number per corncob. The coefficients of variation were below 30%, and the genetic parameters were satisfactory for most of the characteristics. The greatest gains with direct selection were in production (21.23%) and productivity of commercial sweet corn ear (15.19%), however, the indirect gains are unsatisfactory for sweet corn selection, and the sum of ranks index provided a balanced distribution of gains. The sweet corn genotypes L2P11, L2P37, P45, L2L5P3, and L5P18 presented a superior performance for the set of characters evaluated. Key words : Zea mays subsp. saccharata, plant biometrics, yield gain prediction, sum of ranks index Ernane Miranda Lemes () Email: ernanefito@gmail.com The Korean Society of Crop Science