Open access (OA) publishing, whereby authors, their institutions, or their granting bodies pay or provide a repository through which peer-reviewed work is available online for free, is championed as a model to increase the number of citations per paper and disseminate results widely, especially to researchers in developing countries. We compared the number of citations of OA and non-OA papers in six journals and four books published since 2000 to test whether OA increases number of citations overall and increases citations made by authors in developing countries. After controlling for type of paper (e.g., review or research paper), length of paper, authors' citation profiles, number of authors per paper, and whether the author or the publisher released the paper in OA, OA had no statistically significant influence on the overall number of citations per journal paper. Journal papers were cited more frequently if the authors had published highly cited papers previously, were members of large teams of authors, or published relatively long papers, but papers were not cited more frequently if they were published in an OA source. Nevertheless, author-archived OA book chapters accrued up to eight times more citations than chapters in the same book that were not available through OA, perhaps because there is no online abstracting service for book chapters. There was also little evidence that journal papers or book chapters published in OA received more citations from authors in developing countries relative to those journal papers or book chapters not published in OA. For scholarly publications in conservation biology, only book chapters had an OA citation advantage, and OA did not increase the number of citations papers or chapters received from authors in developing countries.