Above-ground biomass estimates are vital for carbon accounting, climate change mitigation planning and other activities. Field-based methods to acquire biomass information are tedious, time consuming and expensive. For this reason, remote sensing methods to estimate above-ground biomass have gained wide attention and use worldwide. Optical sensor data, mainly from Landsat, has been a major data source. However, extensive cloud cover, particularly in tropical forests, has presented difficulties in analyzing and using optical data for estimating biomass. Alternatively, Synthetic Aperture Radar (SAR) has increasingly gained attention as a promising approach to biomass estimation. Few studies have examined how biomass estimation models can be developed based on radar backscatter. Radar sensors that use the microwave region to transmit and receive signals are capable of penetrating clouds making SAR data a viable alternative to biomass estimation, particularly in moist tropical forests such as dipterocarp forests in the Philippines. This paper describes the use of quad-polarimetric images acquired by the Advance Land Observing Satellite (ALOS) Phased Array L-band Synthetic Aperture Radar-2 (PALSAR-2) in conjunction with forest sampling plots to generate spatially explicit above-ground biomass estimates within the Northern Sierra Madre Natural Park (NSMNP) in the northeastern part of the country. Two quad-polarimetric Level 1.1 PALSAR-2 images over NSMNP were pre-processed using ENVI SARscape. Data were collected from 134 forest sampling plots scattered within the study area. The above-ground biomass per hectare (AGB/ha) from these sampling plots was estimated using allometric equations developed for tropical dipterocarp forests. Using the location of the sampling plots, relationships between the radar backscatter in all polarization channels and ground estimated AGB/ha were statistically modeled to identify trends and produce the SAR-derived biomass estimates. The results presented in this paper substantiate the suitability of SAR datasets as an alternative data source that can enhance the conventional approaches of site-level biomass estimation relevant for various national and international reporting requirements.