ABSTRACT A major source of uncertainty in the age determination of old (∼10 Gyr) integrated stellar populations is the presence of hot horizontal branch (HB) stars. Here, we describe a simple approach to tackle this problem, and show the performance of this technique that simultaneously models the age, abundances, and HB properties of integrated stellar populations. For this, we compare the results found during the fits of the integrated spectra of a sample of stellar population benchmarks, against the values obtained from the analysis of their resolved colour–magnitude diagrams (CMDs). We find that the ages derived from our spectral fits for most (26/32) of our targets are within 0.1 dex to their CMDs values. Similarly, for the majority of the targets in our sample we are able to recover successfully the flux contribution from hot HB stars (within ∼0.15 dex for 18/24 targets) and their mean temperature (14/24 targets within $∼ 30 {{\ \rm per\ cent}}$). Finally, we present a diagnostic that can be used to detect spurious solutions in age, that will help identify the few cases when this method fails. These results open a new window for the detailed study of globular clusters beyond the Local Group.