The selection of a survey method of free-roaming dog populations should be based on analyses of local capacities and management priorities. Here, we compare the results of surveys of the stray dog population in Quito, Ecuador, using two different methodologies and propose an alternative method for future surveys in the city. We carried out all surveys in ~5 km-transects in a sample of eight urban and eight rural parishes (16 transects total). In 2018, we used the capture-recapture method to estimate absolute population size and 95% CI. We began transect surveys at 04 h 00 (local time) and identified individuals with photographs. The main limitations of this method were errors in identifying individuals, since photographs were not always clear, partly due to low light conditions during the surveys. This method also required more time and more complex logistics. In 2019, we used distance sampling to estimate population density and began the surveys at 08 h 00 (local time). Errors in the estimation of animal-observer distances and angles were our main concern when using this method. For future surveys, we propose to carry out direct observations of dog abundance (number of free-roaming dogs/km) during street counts, complemented with capture-recapture surveys every 5 years. This alternative method albeit simple, is sensitive enough to (1) provide local authorities with objective assessments of management interventions, (2) better understanding the dynamics of free-roaming dog populations and (3) increasing public awareness about the problem of pet abandonment through citizen participation in the surveys.