CH4 emissions from liquid pig manure stored in under-floor pits were investigated using an empty full-scale pig house in Denmark in winter season. The experiment designs were three manure depths of 0.15, 0.40 and 0.65m, four ventilation rates of 211, 650, 1852 and 3088m3/h, and with or without pit curtains. The three parameters that explained most of the variability of the CH4 emissions were manure temperature, manure depth and ventilation rate. There were no significant differences in CH4 emission rates among four ventilation rates with diffuse ceiling inlet and negative pressure exhaust. However, the lower manure temperature could reduce CH4 emission or compensate the effects of higher ventilation rate. CH4 emission rate reduced with decreasing manure depth in pit. No significant differences in CH4 emission rates were found by using manure pit curtains. The emission rates were more sensitive to manure temperature than to manure depth and ventilation rate due to higher sensitivity percentages. At the lower ventilation rate and manure depth ranges, the emission rates were more sensitive to ventilation rate change compared to the higher ones. However, the contrary sensitivity responses for different manure temperature ranges were obtained.