Building energy efficiency is a key factor in reducing CO2 emissions. For this reason, EU member states have developed thermal regulations to ensure building thermal performance. These results are often based on results achieved with building simulation software during the design stage. However, the actual thermal performance can deviate significantly from the predicted one, and this difference is known as the energy performance gap. Accurate indicators of the actual thermal performance are a valuable tool to guarantee building quality. These indicators, including the heat transfer coefficient (HTC) and the heat loss coefficient (HLC), can be estimated by the application of in situ methods. As multi-family housing and tertiary sector buildings are an important part of the building stock, mature methods to measure their thermal performance are needed. This paper presents a short-duration method for assessing the HTC in large building typologies using a sampling approach. The method was applied in a four-storey building model under different conditions to study the limits of the method and to improve indicator bias and uncertainty. Indicator quality was strongly influenced by the external weather conditions, the temperature variation during the protocol and the heat exchange with the adjacent apartments. Under winter conditions and with stable indoor temperatures, the method had a high accuracy when the protocol was applied for half a day. It is recommended that the protocol be used over two days to improve indicator quality under less favorable test conditions.