In photodynamic therapy (PDT) oxygen plays a vital role in killing tumor cells. Therefore oxygen dosimetry is being thoroughly studied. Light distribution into tissue is modelled for radiation-induced fibrosarcoma (RIF) and nodular basal cell carcinoma (nBCC), in order to study the influence of blood flow on singlet oxygen concentration effectively leading to cell death ([1O2]rx) from PDT, within this light distribution. This is achieved through initial oxygen supply rate (g0) and initial molecular oxygen concentration ([3O2]0) calculations. Monte Carlo simulations and mathematical models are used for spatial and temporal distributions of [1O2]rx. Hypoxia conditions are simulated by minimizing [3O2]0 and g0. Furthermore, an optimization algorithm is developed to calculate minimum initial molecular oxygen concentration needed ([3O2]0,min) for constant [1O2]rx, when blood flow changes. Our results validate that in initially well-oxygenated scenarios with normal blood flow maximum [1O2]rx values are significantly higher than corresponding values of hypoxic scenarios both for RIF and nBCC models, with maximum oxygen supply rate percentage variations being independent from g0. Moreover, [1O2]rx appears to be more affected by an increase of g0 than of [3O2]0 values. For low blood flow there is a linear relationship between [3O2]0,min and g0, while for better oxygenated areas high blood flow reduces [3O2]0,min needed in exponential manner. Blood flow appears to be able to compensate for oxygen consumption. The developed optimization protocol on oxygen dosimetry offers a suitable combination of [3O2]0,min and g0 to achieve constant [1O2]rx, despite possible blood flow variations.