BACKGROUND: In sub-Saharan Africa, house design and ventilation affects the number of malaria mosquito vectors entering houses. This study hypothesized that indoor light from a CDC-light trap, visible from outside a hut, would increase entry of Anopheles arabiensis, an important malaria vector, and examined whether ventilation modifies this effect. METHODS: Four inhabited experimental huts, each situated within a large chamber, were used to assess how light and ventilation affect the number of hut-entering mosquitoes in Tanzania. Each night, 300 female laboratory-reared An. arabiensis were released inside each chamber for 72 nights. Nightly mosquito collections were made using light traps placed indoors. Temperature and carbon dioxide concentrations were measured using data loggers. Treatments and sleepers were rotated between huts using a randomized block design. RESULTS: When indoor light was visible outside the huts, there was an 84% increase in the odds of collecting mosquitoes indoors (Odds ratio, OR = 1.84, 95% confidence intervals, 95% CI 1.74-1.95, p < 0.001) compared with when it was not. Although the odds of collecting mosquitoes in huts with closed eaves (OR = 0.54, 95% CI 0.41-0.72, p < 0.001) was less than those with open eaves, few mosquitoes entered either type of well-ventilated hut. The odds of collecting mosquitoes was 99% less in well-ventilated huts, compared with poorly-ventilated traditional huts (OR = 0.01, 95% CI 0.01-0.03, p < 0.001). In well-ventilated huts, indoor temperatures were 1.3 °C (95% CI 0.9-1.7, p < 0.001) cooler, with lower carbon dioxide (CO2) levels (mean difference = 97 ppm, 77.8-116.2, p < 0.001) than in poorly-ventilated huts. CONCLUSION: Although light visible from outside a hut increased mosquito house entry, good natural ventilation reduces indoor carbon dioxide concentrations, a major mosquito attractant, thereby reducing mosquito-hut entry.