A thorough understanding of carbon storage patterns in forest ecosystems is crucial for forest management to slow the rate of climate change. Here, we explored fine-scale biomass spatial patterns in a secondary warm temperate deciduous broad-leaved forest in north China. A 20-ha plot was established and classified by topographic features into ridge, valley, gentle slope, and steep slope habitats. Total tree biomass varied from 103.8 Mg/ha on the gentle slope habitats to 117.4 Mg/ha on the ridge habitats, with an average biomass of 109.6 Mg/ha across the entire plot. A few species produced the majority of the biomass, with five species contributing 78.4% of the total tree biomass. These five species included Quercus mongolica Fisch. ex Ledeb (41.7 Mg/ha, 38.1%), Betula dahurica Pall. (19.8 Mg/ha, 18.0%), Acer mono Maxim. (12.6 Mg/ha, 11.5%), Betula platyphylla Suk. (7.0 Mg/ha, 6.4%), and Populus davidiana Dode. (4.8 Mg/ha, 4.4%). The five species were also associated with certain habitats; for example, Q. mongolica was positively associated with the ridge habitat and A. mono was positively associated with the valley habitat. Results from this work document the variability in forest biomass across a warm temperate forest ecosystem of north China, with implications for managing and accounting forest carbon.