To provide reference for improving the innovation efficiency of China's pharmaceutical industry and promoting high-quality development of China's pharmaceutical industry.

Data Envelopment Analysis (DEA) was used to measure the innovation (Technology) efficiency under constant scale (CRS) and variable scale (VRS) of pharmaceutical industry. The input and output data of 28 provinces (autonomous regions and municipalities) in China from 2014 to 2018 were selected from China Statistical Yearbook and China High-tech Industry Statistical Yearbook. On this basis, further comparative analysis was conducted from the perspective of industrial cluster to explore the change law of innovation efficiency of China's pharmaceutical industry, and corresponding suggestions were given.

There were great differences in the development of technological innovation efficiency among different regions in China; the innovation input and output of pharmaceutical industry was unbalanced and not directly proportional to the development level of the industry; the innovation efficiency in China was mainly affected by pure technical efficiency. through the analysis, it was found that there is a deviation between the evaluation of innovation efficiency of pharmaceutical industry by using DEA model alone and the actual situation of various regions. We conjectured that the deviation may be caused by the particularities of pharmaceutical industry and the impact of innovation spillover. It is suggested that when using DEA to evaluate the innovation efficiency of pharmaceutical industry, we should consider the factors such as the weak periodicity of pharmaceutical industry, the long output cycle of innovation achievements, and the existence of innovation efficiency spillover, so as to make the results more realistic.