Purpose: The purpose of the study is to analyse the occurrence and distribution of different tree species in Gilgit-Baltistan, Pakistan, as a baseline for further inventories, and estimate the biomass per species and plot. Furthermore, it aims to measure forest biodiversity using established formulae for tree species diversity index, richness, evenness and accumulative curve.Design/methodology/approach: Field data were collected, including stratification of forest sample plots. Statistical analysis of the data was carried out, and locally appropriate allometric equations were applied for biomass estimation.Findings: Representative circular 556 forest sample plots of 1,000 m²contained 13,135 trees belonging to nine tree species with a total aboveground biomass of 12,887 tonnes. Sixty-eight per cent of the trees were found between 2,600 and 3,400 masl; approximately 63 per cent had a diameter at breast height equal to 30 cm, and 45 per cent were less than 12 m in height. The Shannon diversity index was 1.82, and Simpson’s index of diversity was 0.813.Research limitations/implications: Rough terrain, long distances, harsh weather conditions and location of forest in steep narrow valleys presented challenges for the field crews, and meant that fieldwork took longer than planned.Practical implications: Estimating biomass in Gilgit-Baltistan’s forests using locally developed allometric equations will provide transparency in estimates of forest reference levels, National Forest Monitoring System in Pakistan and devising Reducing Emissions from Deforestation and Forest Degradation national strategies and for effective implementation.Originality/value: This paper presents the first detailed forest inventory carried out for the dry temperate and semi-arid cold region of Gilgit-Baltistan, Pakistan.