We use the degree of degradation of crater-associated radar-dark deposits on Venus to estimate the age of the crater and neighboring units. We analyzed craters ≥30 km in diameter superposed on regional plains (subpopulation 1; 138 craters) and on later units (subpopulation 2; 30 craters) and estimated percentages of craters with dark parabolas (DP), clear dark halo (CH), faint dark halo (FH), and no dark halo (NH). We constructed theoretical models of the evolution of these percentages with time for two possible interpretations of the upper boundary of the regional plains: (1) globally synchronous and (2) nonsynchronous (diachronous). They show that in the synchronous case the dark deposit lifetimes are proportional to the corresponding percentages observed for subpopulation 1: TDP = 0.15T, TCH = 0.3T, and TFH = 0.3T, where T is the mean global surface age of Venus. In diachronous cases the percentages may or may not be proportional to the appropriate lifetimes, depending on the range of the regional plains. If the time range of plains emplacement is not larger than ¿0.5T, the use of TDP = 0.15T and TCH = 0.3T is appropriate. We consider the time ~0.5T ago as the lower time boundary for the age of the now-observed CH craters and the time 0.1--0.15T ago as the lower boundary for the age of DP craters. We propose that the Atlian Period of the geologic history of Venus be subdivided into Upper and Lower Epochs, with the boundary between them at ~0.5T ago, and the lower boundary of the Aurelian Period be placed at 0.1--0.15T ago. We apply this approach to assess ages of activity of three volcanic-tectonic structures on Venus: Beta Regio-Devana Chasma (shows evidence of activity younger than 0.5T ago), Mylitta Fluctus (also younger than 0.5T), and Atla Regio (shows evidence of activity younger than 0.1--0.15T ago).