The current popcorn production in Mexico is insufficient to meet the demand for this grain, since there are no genotypes suitable for production within the country. The native popcorn varieties do not meet market standards; therefore, it is necessary to carry out genetic improvement on yield and popping expansion. The objective of this study was to estimate the response to selection on yield and popping expansion of popcorn under two modalities of mass selection. We used a population of popcorn from the first selection cycle (C1) of the cross between a commercial popcorn of the North American Yellow Pearl race and a native population of the Mexican race Palomero Toluqueño. An additional cycle of mass selection (C2) was carried out with stratification (MSS) and without stratification (MNSS) followed by a field trial that included the different cycles of selection. The parents of the initial cross and two controls, under a complete randomized block experimental design with 10 replications in two localities, and the genetic gain per cycle was calculated. Genotypes C2MSS, C2MNSS, C1MSS and the control Palomero Ixtenco showed the highest average grain yield with values of 4.9 t ha−1. For popping expansion, the Jack Superior control showed the best flake volume with 34.7 cm3 g−1, in contrast to the native popcorn Criollo Plaza with a value as low as 7.2 cm3 g−1. Chapingo was the locality with the highest grain yield, averaging 5.08 t ha−1, while in San Salvador Atenco, the yield was 2.78 t ha−1. Genetic gains were found with a popping expansion of 12.50% with stratification and 11.42% without stratification. For grain yield, a gain of 6.82% was obtained using stratification and 0.74% without stratification. Stratified visual mass selection is an efficient method for genetic advancement in popcorn yield, while popping mass selection is efficient regardless of stratification.