Departing from the embodiment assumption that our conceptualizations are grounded in the physical world, gestures should also refer to those conceptualizations (BARSALOU, 1999; CIENKI, 1998A, 1998B, 2008, 2013; GLENBERG & ROBERTSON, 2000; HOSTETTER & ALIBALI, 2008). In this study, we investigate how our conceptualizations of the physical and of the abstract are expressed in speech and gesture, using the four-way spatial distinction found in Brazilian Portuguese between ‘aqui’ (near to speaker), ‘aí’ (near to addressee), ‘ali’ (near to both speaker and addressee), and ‘lá’ (distant to both). We tested two opposing hypotheses: 1) that gestures used with concrete and abstract deixis may be similar to each other, based on claims from embodiment theory, and 2) that gesture use may differ in concrete and abstract deixis, based on claims from neuroscience and based on patterns of usage of these deictic words found in a corpus of spoken Brazilian Portuguese. Twenty-four participants were asked to act out small scripts with eight contexts, each containing one occurrence of both concrete and abstract uses of ‘aqui’, ‘aí’, ‘ali’, and ‘lá’. The results show the semantic opposition between 'aqui' and 'lá' is also present in co-verbal gesture. But there was not a clear difference in gesture use with ‘aí’ as compared with the other key words, as one might have anticipated from the use of the word in the C-ORAL Brasil corpus. Gestures with concrete use of the key words are similar in some ways to gestures with abstract use, but there are also many differences. In conclusion, imagery seems to be activated with abstract reference using these spatial adverbial pronouns, but the factors motivating the particularities of the differences remain to be explored in future work.