Biomass pyrolysis is an efficient and economical conversion process. The kinetics of biomass conversion is still not fully understood mainly because this complex material is composed of numerous compounds with very different reactivities. In this work an inverse model is used to determine the nonparametric probability density function p(Ea) of activation energy (Ea) for coconut shell waste (CSW) pyrolysis via the Ramped PyrOx (RPO) method from the time-series and temperature-series mass conversion data. In this method, the degradation rates are described by inverting distributed activation energy model (DAEM). This method does not require p(Ea) to follow any particular parametric form furthermore the modelling results are independent of experimental conditions such as heating rates. The pyrolytic kinetics is modelled considering first-order degradation mechanism. The activation energies of pseudo-components were estimated as 246.5 kJ∙mol⁻¹, 264.2 kJ∙mol⁻¹, 290.4 kJ∙mol⁻¹ and 319.7 kJ∙mol⁻¹.