The marine ecological models have been widely applied. The key factor limiting application of ecological models is the uncertainty of these models, primarily due to the uncertainty of the parameters used. The purpose for sensitivity analysis of parameters for ecological models was to assess influential degrees of various parameters on simulating outcomes with a particular model, which are the essential procedures for parameter optimization and model calibration and the important tools for understanding the behaviors of the model. In this study, we conducted sensitivity analysis on 50 parameters for five state variables i.e. phytoplankton, zooplankton, nutrient, detritus and dissolved oxygen included in the coupled biological-physical model for Jiaozhou Bay. We found that among 50 parameters, three were highly sensitive, two were sensitive, 11 were moderately sensitive and 34 were less sensitive. The highly sensitive and sensitive parameters included phytoplankton growth rate (μ_PRPC), factors for dark reaction correction (FAC), light saturation intensity (α), phytoplankton death rate (μ_DEPC) and light extinction coefficient of water (bla). They have major impacts on plant growth and death, reflecting the essentialness and importance of phytoplankton in ecosystem. The plant growth process has the most important impacts on the stimulating results. The key factor limiting plant growth is light whereas light extinction of water is the most important factor limiting light intensity. These five parameters have significant impacts on carbon and nutrient cycles, and are the most important parameters in Jiaozhou Bay ecosystem. Thus, they should be optimized with the first priority. The impacts of moderately sensitive parameters are mediated mainly via the influence of nutrient on phytoplankton growth and death, temperature influence on light saturation intensity and the influence on zooplankton growth, grazing and death as well as the impacts of phytoplankton biomass on grazing, chlorophyll a generation, phosphorus release from disposed substances under hypoxia condition and phosphorus absorption by plants. There are different degrees of variations in sensitivity of the state variables and thus, they have different characteristics. The less sensitive parameters-related processes are mainly light extinction of chlorophyll a and detritus, temperature limitation to zooplankton growth, zooplankton grazing, and mineralization of detritus and sediment, mineralization/deposition of detritus and sediment, the inorganic nitrogen-related processes and changes in dissolved oxygen concentration. These processes are influenced not only by model's internal parameters but also by external factors such as water depth, sea water temperature and pollutions from land-based sources. Moderately and less sensitive parameters affect model's local processes, and thus are the important basis for model calibration. Additionally, four state variables i.e. chlorophyll a, zooplankton, detritus and inorganic phosphorus, are calibrated according to the maximal production constant of chlorophyll a (KCHmax), first-order zooplankton death rate (μ_DEZC1), mineralization rate of organic detritus (μ_REDC) and half-saturation constant of phosphorus absorption by zooplankton (hUPPP). The sensitivity analysis of nutrient-related parameters has shown that phytoplankton in Jiaozhou Bay is limited by phosphorus and however inorganic nitrogen is principally effected by pollutions from land-based sources. Dissolved oxygen is less sensitive on each parameter.