Photovoltaic power generation (PPG) capacity credit is an important index for evaluating the contributions of photovoltaic power on system reliability. A novel calculation method for PPG capacity credit is presented considering the correlation between output of photovoltaic power and load. The method firstly presents a time-interval based output model of PPG for day-mode and night-mode, then constructs a joint multi-state model, which can be used to characterize the correlation between the output of PPG and load, using a discrete joint probability distribution, and finally calculates the system reliability indices and capacity credit of PPG using the non-sequential Monte Carlo simulation method. Based on equivalent LOLE and LOEE criterion, the capacity credit of the RBTS test system containing PPG is analyzed. The results indicate the validity and practicability of the proposed model.