AbstractThe typical approach to modeling building degradation is based on consecutive visual inspections using generic deterioration ratings to characterize the conditions of elements. Our research used snapshot data to capture a range of ages efficiently, and very clearly defined element-specific deficiency-based conditions to reduce subjectivity. Snapshot data were obtained of 12 distinct building elements, from buildings owned by 7 local councils in Sri Lanka (with ages up to 60 years), using both generic deterioration– and deficiency-based condition ratings. The deficiency-based ratings were found to be lower than the generic deterioration–based ratings at early ages but higher at later ages. Two data-driven indicators were used to estimate the relatively maintenance-free life of each element. Markov degradation models were developed for both types of ratings. The deficiency-based predictions were found to be more accurate than generic deterioration–based ones. Slabs, beams, and columns had the lowest rate of degradation; timber doors, timber windows, ceilings, wall plaster and floor tiles had higher rates; and ceiling fans, fan regulators, wall paint, and rendered cement floors had the highest rates.