Central in the frameworks proposed to capture the knowledge needed for teaching mathematics is the assumption that teachers need more than pure subject-matter knowledge. Validation studies exploring this assumption by recruiting contrasting populations are relatively scarce. Drawing on a sample of 644 Greek-Cypriots preservice and inservice elementary school teachers and university students with strong mathematical background, this study explored participants' knowledge with respect to four teaching practices: providing and evaluating explanations; selecting and using representations; analyzing student errors, misconceptions, and non-conventional solutions; and selecting tasks. Results emerging from an item response theory (IRT) model and non-parametric tests showed significant differences among these groups, largely in the first two practices. Clinical interviews with a voluntary sub-sample of these participants provided insights into the unexpected findings related to the other two practices. The theoretical, methodological, and practical implications of these findings are discussed, and directions for future studies are offered.