This article reports on two professional development programs for mathematics teachers and their effects on student achievement. The projects' design and their implementation within a larger municipality in Sweden, working together with over 90 teachers and 5000 students in elementary school, are described by using a set of core critical features of teacher professional development as identified in the research literature. Data on student achievement were collected at the beginning of the projects as well as one year after their implementation. The results demonstrate a significant interaction effect between year and intervention for grades 2, 8, and 9. Upon a closer examination of the descriptive statistics, the first project points at slight improvements for students in grade 2, while the results for grades 8 and 9 indicate a declining trend. The results for the second project are markedly different and indicate an improvement for student achievement in grades 8 and 9, while the results for grade 2 suggest a deterioration. Considering that both projects are conducted within the same context and are very similar when it comes to several of the critical features of professional development as identified in the research literature, these variances in results are noteworthy. Therefore, building on a closer examination of the projects' critical features, the Swedish educational context, and results from teacher evaluations, the article ends with a discussion of possible explanations for the variations found between, as well as within, the two projects' differing impact on student achievement.