The forward-looking nature of monetary policy formulation under an inflation targeting (IT) regime requires the accurate and timely assessment of key macroeconomic variables, such as inflation. This study develops nowcasting models for headline and core inflation in the Philippines, in the form of time-varying regressions using state-space representation. To exploit the information content of more available high-frequency price and financial data, the statespace models are analyzed using the Kalman filter, which updates estimates upon the availability of new information. Based on one-month-ahead forecasts, the forecasting performance of the state-space models has been found to be at par with that of existing linear inflation forecasting models at the BSP. In addition, the paper serves as an initial exercise for modeling structural changes in inflation.