Previous research found that bilinguals outperform monolinguals on cognitive control tasks, but few studies have examined differences within the bilingual population. This study aims at comparing cognitive control performance in groups of bilinguals divided along their proficiencies in two languages. Spanish-English adult bilinguals (n=47) performed a nonverbal rule-switching task while inside the fMRI scanner. Participants were required to respond as quickly as they could using a button box to either the color or shape of stimuli, with the dimension of interest changing (switch trials) or staying the same (non-switch trials), indicated by a cue. Proficiency was measured using the vocabulary and sentence comprehension portions of the Woodcock Language Proficiency Battery-Revised in English and Spanish. Balanced bilinguals were defined as those individuals who have similar scores on both the English and Spanish tests. Unbalanced bilinguals are those with large differences between their English and Spanish proficiencies. Our results indicate that proficiency balance has significant effects on neural activation while bilinguals engage in a nonverbal rule-switching task. Specifically, unbalanced bilinguals present with large activation patterns in brain areas related to executive function in the rule-switch condition. In contrast, there are little to no activations in the unbalanced group in the non-switch condition, and under both conditions in the balanced group. From these results we can conclude that nonverbal task-switching is more effortful for the unbalanced group than the balanced group. These results are consistent with the view that the bilingual advantage most likely emerges from the consistent use of two languages.