Population-level studies often require age estimation of fish, but populations in small rivers and streams are generally smaller than those in large rivers or reservoirs. Therefore, non-lethal aging methods are generally recommended to minimize the potentially negative effects of sampling on population size. Accordingly, our main goal was to compare otoliths and scales as structures for estimating the age of redbreast sunfish (Lepomis auri- tus) and green sunfish (L. cyanellus) in an urban watershed. Reader agreement was greater for otoliths (88%–89%) than for scales (73%–79%), and pre- cision (mean CV) in age estimates was better for otoliths (3.7%–4.0%) than scales (6.1%–9.4%) for both species. Readers were significantly more confi- dent in their otolith-derived age estimates than scale-derived age estimates for both species. For redbreast sunfish and green sunfish, age estimation bias between readers was apparent for scales, but not for otoliths. Otolith-estimated ages of redbreast sunfish ranged from 0 to 7 years, while scale-estimated ages ranged from 1 to 6 years. For green sunfish, otolith-estimated ages ranged from 1 to 7 years, while scale-estimated ages ranged from 0 to 5 years. Mean estimated ages were significantly different between otolith-based and scale-based ages for both species. Scales generally overestimated otolith- assigned ages of younger redbreast sunfish and underestimated otolith-assigned ages of older fish. For green sunfish, scales generally underestimated otolith-assigned ages of older fish. Annual survival estimates derived from catch-curve analyses also differed between otoliths and scales for both spe- cies. We recommend that biologists avoid using scales to estimate the ages of redbreast sunfish and green sunfish. When a non-lethal aging method is required, biologists should use extreme caution in using scales, and explore other non-lethal methods for age estimation.