This study uses an extensive sample of company data referring to seven Latin American countries to analyze the potential influence of simultaneity and heterogeneity in determining decisions to commit to R&D and innovation outputs. First, the analysis focuses on the decisions about spending on in-house and external research. Second, the outcomes of three different innovations: process, product, and organizational. Choices were considered simultaneously by employing a multivariate probit estimation model. In so doing, the enquiry considered the possible systematic interdependencies among the decisions. The results indicate that the two R&D decisions are interdependent. However, a different picture emerged in relation to innovation outputs, in which the correlations coefficients were small with unexpected signs. This may indicate a structural weakness of the technological production system in Latin America, wherein the synergies, complementarities, and complex relationships usually involved in the process of innovation are not fully developed.