Two-phase flow in vertical wells is a common occurrence in oil and gas production. Heavy oil upward vertical flow in wells presents a new challenge for predicting pressure gradient and liquid holdup due to lack of understanding flow behavior and specifically predicting flow pattern in vertical wells. Current two-phase flow mechanistic models have been validated and tuned based on low viscosity data for which they show accurate predictions. The objective of this study is to theoretically investigate flow pattern for high viscosity liquid conditions two-phase flow in vertical pipes. High liquid viscosity flow pattern two-phase flow data will be collected in this study from open literature against which existing flow pattern models will be evaluated to identify modeling discrepancy and gaps. A theoretical modeling of the identified flow patterns will be carried out to develop new models and/or modify the existing ones. The developed models will then be incorporated in Ansari et al. (1994) comprehensive mechanistic model and Zhang et al. (2003) unified mechanistic model to predict pressure gradient and liquid holdup. Finally, the prediction of mechanistic models will be validated with high viscosity experimental data and compared with other existing low viscosity comprehensive models to show the improvements of proposed flow pattern models on pressure gradient and liquid holdup.