GC is commonly used for the analysis of cannabis samples, e.g. in forensic chemistry. However, as this method is based on heating of the sample, acidic forms of cannabinoids are decarboxylated into their neutral counterparts. Conversely, HPLC permits the determination of the original composition of plant cannabinoids by direct analysis. Several HPLC methods have been described in the literature, but most of them failed to separate efficiently all the cannabinoids or were not validated according to general guidelines. By use of an innovative methodology for modelling chromatographic responses, a simple and accurate HPLC/DAD method was developed for the quantification of major neutral and acidic cannabinoids present in cannabis plant material: Delta9-tetrahydrocannabinol (THC), THC acid (THCA), cannabidiol (CBD), CBD acid (CBDA), cannabigerol (CBG), CBG acid (CBGA) and cannabinol (CBN). Delta8-Tetrahydrocannabinol (Delta8-THC) was determined qualitatively. Following the practice of design of experiments, predictive multilinear models were developed and used in order to find optimal chromatographic analytical conditions. The method was validated following an approach using accuracy profiles based on beta-expectation tolerance intervals for the total error measurement, and assessing the measurements uncertainty. This analytical method can be used for diverse applications, e.g. plant phenotype determination, evaluation of psychoactive potency and control of material quality.