This paper investigated the variation of building heating energy consumption caused by global warming in Tianjin, China. Based on the hourly historical and monthly projected future (B1/A1B emissions scenarios) meteorological data, the variation of those relevant meteorological parameters was first analyzed. A TRNSYS simulation model for a reference building was introduced to investigate historical variation of office building energy consumption. The results showed that the 10-year-average heating energy consumption of 2001–2010 had reduced by 16.1% compared to that of 1961–1970. By conducting principal component analysis and regression analysis, future variation of building heating load was studied. For B1/A1B emissions scenarios, the multi-year-average heating load was found to decrease by 9.7% (18.1%)/10.2% (22.7%) compared to that of 1971–2010 by 2011–2050 (2051–2100).