Developing an accurate water level monitoring system is one of the measures to mitigate the effects of water-related hazards such as river flooding. While current monitoring systems in the country are efficient in terms of accurate and immediate data delivery, these systems can be costly. This study assesses the Global Navigation Satellite System (GNSS) performance of low-cost receiver systems for water level monitoring using real-time kinematic (RTK) solution. A total of 10 days’ valid observation were analyzed to compare the two base-rover receiver setups: 1) low-cost base to low-cost rover (LC-LC) and 2) survey-grade base to low-cost rover (SG-LC) grounded on accuracy, integrity, continuity, availability, and cost. Accuracy results show LC-LC=5.81 cm and SG-LC=5.37 cm mean difference of RTK from in-situ readings. In terms of RTK and post-processing kinematic (PPK) difference for integrity criterion, the RTK SG-LC setup has a lower range of RMS of 0.86 to 1.94 cm versus LC-LC setup of 1.19 to 2.28 cm. For the continuity criterion, the average fixed solutions percentage for the LC-LC setup: RTK=91.43%, PPK=92.92%, whereas for the SG-LC: RTK=95.51%, PPK=98.39%. On availability, the number of valid satellites (NSat) and position dilution of precision (PDOP) of RTK and PPK solutions for each setup are LC-LC: RTK=11, PPK=23, PDOP=1.0 and SG-LC: RTK=11, PPK=24, PDOP=1.9. Lastly, in terms of costing, LC-LC costs Php 58,340 while SG-LC costs Php 1,279,645. Overall, the parity of LC-LC with SG-LC in terms of the five criteria suggests viability of using LC-LC for accurate real-time water level monitoring.