This study aims to compare the performance of planted and unplanted constructed wetlands with gravel- and raw rice husk-based media for phenol and nitrogen removal. Four laboratory-scale horizontal subsurface-flow constructed wetland units, two of which planted with cattail (Typha latifolia) were operated outdoors. The units were operated at a nominal hydraulic retention time of 7 days and fed with domestic wastewater spiked with phenol concentration at 300 mg/L for 74 days and then at 500 mg/L for 198 days. The results show that planted wetland units performed better than the unplanted ones in the removal and mineralization of phenol. This was explained by the creation of more micro-aerobic zones in the root zone of the wetland plants which allow a faster rate of phenol biodegradation, and the phenol uptake by plants. The better performance of the rice husk-based planted wetland compared to that of the gravel-based planted wetland in phenol removal could be explained by the observation that more rhizomes were established in the rice husk-based wetland unit thus creating more micro-aerobic zones for phenol degradation. The role of rice husk as an adsorbent in phenol removal was considered not of importance.