In this study, natural brucite was employed as a magnesium source to replace conventional dead-burned MgO for the preparation of magnesium ammonium phosphate cement (MAPC) and magnesium potassium phosphate cement (MKPC) through reactions with ammonium dihydrogen phosphate and potassium dihydrogen phosphate, respectively. The influence of phosphorus-to-magnesium (P/M) molar ratio on the workability, mechanical properties, phase assemblages, and microstructure of both cement systems was systematically investigated. The experimental results indicated that, with decreasing P/M molar ratio, the fluidity of both pastes exhibited a gradual decline, whereas the setting time showed opposite tendencies, i.e., it shortened for MAPC but prolonged for MKPC. The compressive strength of both systems initially increased and then decreased, with the optimal comprehensive performance achieved at a P/M molar ratio of 1/4 for both cements. At this optimal ratio, the 1 d and 28 d compressive strengths of MAPC reached 28.8 MPa and 52.5 MPa, respectively, whereas those of MKPC were 20.0 MPa and 43.7 MPa, respectively. Microstructural characterization revealed that the primary hydration products of MAPC were MgNH4PO4·6H2O and calcium phosphate compounds, the formation of which decreased with decreasing P/M ratio due to the insufficiency of ammonium dihydrogen phosphate in the system. In contrast, the hydration products of MKPC were identified as MgKPO4·6H2O and calcium phosphate compounds, with unreacted brucite and its associated mineral phases still detectable after 28 days of hydration. Additionally, both cement systems exhibited increased structural compactness and reduced total porosity at lower P/M ratios. Collectively, these findings demonstrate that natural brucite shows promising potential as a sustainable alternative to dead-burned MgO for producing high-performance magnesium phosphate cements, although marked differences in reaction kinetics and hydration behavior were observed between the ammonium and potassium systems.