In a previous field-based study, how old one looks for one’s age (perceived age) was found to be predictive of mortality in elderly individuals. In conjunction, perceived age is of relevance and interest to the layperson. Here, a clinical methodology for generating perceived age as a biomarker of facial ageing is detailed. The methodology utilises facial photographs of subjects to present images to large numbers of age assessors who are primarily nationals of the country of study origin. In five observational studies in five different countries involving 874 female subjects it was found that subject age and assessor gender, nationality, age and ageing expertise had little effect on the perceived age data generated. However, increasing the numbers of age assessors up to 50 substantially increased the reproducibility of the mean perceived age for an image and a minimum of 10 assessors were required to give reproducible data. This methodology was also compared to a methodology that utilises passport-type photographs of subjects typically taken in field studies. Although the perceived age data from the two types of images were more similar to each other than to chronological age, there was a marked difference between the two sets of data. Therefore, to allow meaningful comparisons across perceived age studies, the same type of image should be used for the generation of perceived age. In conclusion, the methodology detailed here has demonstrated that perceived age can be a reproducible measure when large numbers of adult age assessors are used and can be utilised globally in studies to investigate facial ageing.