The analysis of spent filters from human-smoked (HS) cigarettes has been used to estimate cigarette yields for over three decades. Until recently, the whole filter was used for estimation; however a part-filter method has been shown to improve the accuracy of estimated HS yields. The part-filter method uses only the mouth-end portion of the filter, downstream of the ventilation holes, for analysis. In this portion, the filtration efficiency is relatively constant irrespective of typical puff flow rates of humans and also minimizes butt length effects (e.g. nicotine condensation) on filtration efficiency. Therefore, the estimations of HS cigarette yields are more robust to human smoking conditions than previous whole-filter methods. British American Tobacco has adopted this method to obtain better understanding of how smokers actually use their products in their everyday environment. This can give information to help understand approaches to harm reduction. Since adopting this method, modifications and quality control features have been added to improve the accuracy of the estimation. This paper will describe in detail the methodology currently in use, along with sources of error, storage studies, quality control, repeatability and reproducibility.