Beginning January 2012, a humane method of dog population management using a Catch-Neuter-Vaccinate-Release (CNVR) program was implemented in Dhaka City, Bangladesh as part of the national rabies control program. To enable this program, the size and distribution of the free-roaming dog population needed to be estimated. We present the results of a dog population survey and a pilot assessment of the CNVR program coverage in Dhaka City. Free-roaming dog population surveys were undertaken in 18 wards of Dhaka City on consecutive days using mark-resight methods. Data was analyzed using Lincoln-Petersen index-Chapman correction methods. The CNVR program was assessed over the two years (2012-2013) whilst the coverage of the CNVR program was assessed by estimating the proportion of dogs that were ear-notched (processed dogs) via dog population surveys. The free-roaming dog population was estimated to be 1,242 (95 % CI: 1205-1278) in the 18 sampled wards and 18,585 dogs in Dhaka City (52 dogs/km2) with an estimated human-to-free-roaming dog ratio of 828:1. During the two year CNVR program, a total of 6,665 dogs (3,357 male and 3,308 female) were neutered and vaccinated against rabies in 29 of the 92 city wards. A pilot population survey indicated a mean CNVR coverage of 60.6% (range 19.2-79.3%) with only eight wards achieving > 70% coverage. Given that the coverage in many neighborhoods was below the WHO-recommended threshold level of 70% for rabies eradications and since the CNVR program takes considerable time to implement throughout the entire Dhaka City area, a mass dog vaccination program in the non-CNVR coverage area is recommended to create herd immunity. The findings from this study are expected to guide dog population management and the rabies control program in Dhaka City and elsewhere in Bangladesh.