The franciscana dolphin has been considered the most threatened small cetacean in the south-western Atlantic Ocean due to gillnet by-catch. The estimation of the species' abundance has been recommended as the highest research priority. A line transect aerial survey to estimate franciscana abundance in Rio Grande do Sul, southern Brazil, was carried out in February 2004. The overall surveyed area comprised 13,341 km2 and at least 20 transect lines. Abundance was estimated using distance sampling methods and assuming g(0) = 0.304. The corrected density is 0.51 franciscanas/km2, resulting in an abundance estimation of 6839 franciscanas (CV = 32%; 95% CI = 3709–12,594) for the surveyed area in Rio Grande do Sul. To improve this estimate: (a) perception bias should be determined; (b) the parameters influencing availability bias should be identified and quantified; and (c) survey sample size should be increased. While the lack of data to correct for perception bias and group size underestimation in this aerial survey is likely to yield an underestimate of franciscana abundance, the use of surfacing and diving time data from boat and land-based surveys to correct for availability bias is likely to cause its overestimation. Alternative values of the g0 group-size estimates and rates of increase were incorporated in the analyses, creating 240 different estimates of annual increment for this franciscana population. Even in the most optimistic scenario, the annual increment of franciscanas is not sustainable with the current levels of by-catch in Rio Grande do Sul, and fishery management to reduce by-catch must be initiated promptly.