OBJECTIVES: The goal of this study was to examine the impact of physical pain on depression using longitudinal survey data for general population in the United States. METHODS: This work employed two rounds of Medical Expenditure Panel Survey (MEPS) from years 2008 and 2009. Depression was measured by frequency of feeling depressed over the last 2 weeks, scaled by 0-not at all, 1-several days, 2- more than half the days, and 3-nearly every day.  Physical pain was measured by severity of pain scaled by 1-not at all, 2-a little bit, 3- moderately, 4-quite a bit, and 5-extremely. People older than 18, who had reported severity of depressed mood and recent physical pain, marriage status, family size, and highest education degrees were included in the study. Only round 2 and round 4 of the survey were used since pain questions were only asked in these two rounds. The final panel contained 21,257 observations, among which 46.32% and 45.34% reported pain limited normal work in round 2 and round 4 respectively; 28.25% and 27.26% documented depressed mood in round 2 and 4 respectively. Ordinary Least Squares (OLS), Linear Mixed Effect Model (LME), generalized linear model (GLM) were used to examine the impact of pain on depression. RESULTS: Compared with GLM and LME, the OLS estimates were shown upward biased. GLM and LME both suggested that individuals whose physical pain deteriorated to the next level from round 2 to round 4 would present a 0.16 (P<0.0001) more depressed mood (based on 0-3 scale) on average. Individuals perceived better health status, were older, richer, married, and employed were less depressed.   CONCLUSIONS: This work utilized a national representative longitudinal data to examine the impact of physical pain on depression. Severity of pain and some individual characteristics were found significantly affecting the severity of depression.