Evidence of a fabric performance gap has underlined the need for measurements of in situ building performance. Steady state co-heating tests have been used since the 1980s to measure whole building heat transfer coefficients, but are often cited as impractical due to their 2-4 week test duration and limited testing season. Despite this, the required conditions for testing and test duration have never been fully assessed. Analysis of field tests show that in 12 of 16 cases, a heat loss estimate to within 10% of the result achieved across a full test period can be achieved within just 72 h. These results are supported by simulated tests upon a wider range of dwellings and across wider environmental conditions. However, systematic errors may still exist, even in cases of convergence and cases with significant uncertainties may never converge. Simulated examples of traditional dwellings and those built in line with current building regulation limits may be tested for more than half the year. However, even when simulated with reduced uncertainties, dwellings with low heat loss and high solar gains, such Passivhaus dwellings and apartments, could be successfully tested for just 22% and 12% of a year respectively, demonstrating the limitations of the co-heating method in assessing such dwellings.