For Spanish nouns, masculine gender is unmarked and feminine is marked. Effects of markedness on gender agreement processing are inconsistent, possibly owing to differences between online methods. This study presents a reanalysis of eye-tracking data from Keating's (2022) study on the processing of noun-adjective gender agreement in speakers of Spanish as a majority and a heritage language. Pairwise comparisons by noun class showed that both groups displayed sensitivity to gender violations with masculine nouns earlier in the time course of agreement processing than they did to violations with feminine nouns, although sensitivity to errors with both classes emerged later for heritage speakers. The time course advantage for detecting gender violations with masculine nouns disappeared when a word occurred between nouns and adjectives. In sum, the native Spanish parser utilizes markedness information during online computation of gender agreement dependencies, even when exposure to native language input is significantly reduced.