Laying-hen housing design and management are the most significant factors affecting the generation and release of gaseous ammonia to the atmosphere. Transitioning the hen housing type from traditional high-rise (where manure is stored within the house for about one year) to modern manure-belt style (where manure is removed every 1 to 4 d and placed into long-term storage) has significantly improved in-barn air quality and reduced farm-level ammonia emissions. As a direct result of the advantages, 100% of new construction for U.S. egg production incorporates manure-belt systems that regularly remove manure from the houses. However, manure-belt system designs (e.g., active vs. passive drying of manure on the belt) and management practices (e.g., frequency of manure removal) vary considerably across the industry, leading to large variations in system performance and efficiency. Thus, questions remain about the optimal design and management of manure-belt facilities to achieve the desired reductions in ammonia emissions. As part of the Coalition for a Sustainable Egg Supply (CSES) project, 27 months of continually monitored environmental data (including ammonia and greenhouse gas emissions) were collected from three hen-housing systems: a conventional cage house (CC) with a 200,000-hen capacity, an enriched colony house (EC) with a 50,000-hen capacity, and an aviary house (AV) with a 50,000-hen capacity. All three hen houses were located on the same farm and were populated with Lohmann white hens of the same age. All houses were equipped with manure-drying air ducts above the manure belts using recirculated indoor air (flow rate ranging from 0.46 to 1.49 m³ h^-1 hen^-1). Manure on the belts was completely removed every 3 to 4 d. Average daily house-level ammonia (NH₃) and carbon dioxide (CO₂) emissions as affected by manure accumulation time (MAT, from 1 to 4 d) on the manure belts were analyzed. Results indicate that for all three types of houses, NH₃ emission rates (g hen^-1 d^-1) were significantly lower for MAT of 1 and 2 d (mean ±SE of 0.061 ±0.005 and 0.064 ±0.004, respectively) than for MAT of 3 and 4 d (0.085 ±0.005 and 0.115 ±0.007, respectively) (p < 0.001). Emissions of CO₂ (g hen^-1 d^-1) were not significantly affected by MAT, averaging 67.8 ±5.7 for CC, 74.7 ±10.2 for EC, and 75.9 ±10.5 for AV. Estimating annual NH₃ emissions from each type of house revealed that shortening the manure removal interval from every 4 d to every 2 d has the potential of reducing NH₃ emissions by 27% for the CC and EC houses and by 19% for the AV house. However, verification of the potential reductions is needed.