Bark of Daphne bholua is an important non-timber forest product and makes a substantial contribution to the Nepalese economy. A precise estimate of the amount of D. bholua bark in mountain forests is possible using a biomass model. We developed an allometric bark biomass model for naturally grown D. bholua in Baglung District in the mid-hills of Nepal. The model was based on data from 101 destructively sampled D. bholua on 20 sample plots representing different growth stages (regeneration, established, and matured), site qualities, and stand densities, and we used diameter and height–diameter ratio as predictors. Among 9 functions evaluated, a simple power function showed the best fit to the data. This model described most of the variations in bark biomass with no substantial trends in the residuals. Leave-one-out cross-validation also confirmed the high precision of this model, because it described most of the variations in bark biomass with no substantial trends in the prediction errors. The model can be applied for a precise prediction of bark biomass for individuals of D. bholua with diameters and height–diameter ratios similar to those used in this study. It is site-specific, and its application should therefore be limited to sites with growth stage, site quality, stand density, and species distribution similar to those that formed the basis of this study. Further validation and verification of this model, with a larger dataset collected from sites with a wider range of these characteristics, is recommended.