The aim of this study was to determine the effect of problem-based learning model and visual-spatial intelligence in developing student’s geometry achievement. This study used non-equivalent control group design and the sample collection technique was purposive sampling. The populations were students of class VIII SMP 7 Padang in the school year of 2017/2018 involving 32 students of class VIII.2 as an experimental class and 32 students of class VIII. 1 as the control class. The instrument used in this study was test of visual-spatial intelligence and geometry achievement. The data were analyzed quantitatively by using independent samples two path analysis of variance. Based on the data analysis, it was found that: (i) the geometry learning outcomes of students learning with PBL model was higher that those learning with DL model; (2) geometry learning outcomes of students having high spatial-visual intelligence were higher than students having low spatial-visual intelligence; (iii) there was interaction between the application of PBL model and spatial-visual intelligence; (4) students having high spatial- visual intelligence got higher geometry learning outcomes after learning with PBL model compared with that after learning with DL model.