Substance use is a serious public health concern. Despite advances in the theoretical conceptualization of within and between person pathways to substance use, researchers are limited by the longitudinal models currently available. Researchers often fit linear mixed effects models (L-MM) and proportional odds mixed effects models (PO-MM) to longitudinal ordinal data with many response categories defined by collapsed count data (e.g., 0 drinking days, 1-2 days, 3-6 days, etc.). Consequently, existing models ignore the underlying count process, resulting in a disjoint between the construct of interest and the models being fitted. My proposed novel ordinal-count mixed effects modeling framework overcomes this limitation by explicitly linking ordinal responses to a suitable underlying count distribution. In doing so, researchers can fit ordinal negative binomial mixed effects models (ONB-MM) and ordinal zero-inflated negative binomial mixed effects models (OZINB-MM) to ordered data as if they had directly observed the underlying discrete counts. The utility the ONB-MM and OZINB-MM was verified by simulation studies. The simulation studies demonstrated that the proposed ordinal-count models recovered the underlying unobserved count process across a range of conditions that may arise in the study of substance use. Results also showed the advantages of the proposed ordinal-count mixed effects models compared to existing L-MM and PO-MM. In sum, my ordinal-count mixed effects modeling framework offers several quantitative and substantive advantages over currently available methods.