Liquid dairy manure storage is favourable for methane (CH4) production and a significant source of this greenhouse gas. Manure storage in cold climates faces large temperature variations over the course of a year, and added on-farm complexities in manure loading require year-round measurements which are currently lacking. The objectives of this study were to 1) quantify CH4 emissions from farm-scale liquid dairy manure storage in a cold climate, 2) investigate the effect of manure temperature on CH4 emissions, and 3) compare measured CH4 emissions with values derived from the USEPA temperature model. Methane fluxes were measured from Aug 2010 to early Nov 2011 on a commercial farm in Ontario, Canada, using a micrometeorological mass balance method which relied on 3 towers with air intakes at 4 heights placed around the storage tank. Manure temperature and volatile solid content (VS) were also measured. Monthly CH4 emissions scaled by VS decreased from 43.8 g CH4 kg−1 VS in Sep 2010 to 5.3 g CH4 kg−1 VS in Jan 2011, and were correlated with temperature (r2 = 0.94). Manure temperature increased starting in Feb 2011, but the increase in CH4 emissions was delayed to Jul 2011. Hence, the response of CH4 emissions to temperature showed a hysteretic pattern where emissions in the warming branch after the winter (Apr-Jun) were significantly lower than emissions in the cooling branch after the summer (Sep-Nov) despite similar temperatures. The temperature model predicted emissions during the cooling phase well, but overpredicted emissions during the warming phase, indicating other limiting factors for microbial CH4 production. Thus, the annual predicted CH4 emissions were about 2.3× larger than the measured values. The newly observed hysteresis effect should be considered when predicting CH4 emissions from stored manure.