Background Studies comparing novel collection methods for host seeking and resting mosquitoes A. arabiensis were undertaken in a village in Eritrea. Techniques included an odor baited trap, a novel tent-trap, human landing collection and three methods of resting collection. A technique for the collection of mosquitoes exiting vegetation is also described. Pre-gravid rates were determined by dissection of host seeking insects and post-prandial egg development among insects collected resting. Results Overall 5,382 host-seeking, 2,296 resting and 357 A. arabiensis exiting vegetation were collected. The Furvela tent-trap was the most efficient, risk-free method for the collection of outdoor host-seeking insects, whilst the Suna trap was the least effective method. Mechanical aspirators (the CDC backpack or the Prokopack aspirator) were superior to manual aspiration in a dark shelter but there was no advantage over manual aspiration in a well-lit one. An estimated two-thirds of newly-emerged mosquitoes went through a pre-gravid phase, feeding twice before producing eggs. Mosquitoes completed gonotrophic development in a dark shelter but left a well-lit shelter soon after feeding. One blood-fed female marked in the village was recaptured 2 days after release exiting vegetation close to the oviposition site and another, shortly after oviposition, attempting to feed on a human host 3 days after release. Exit rates of males from vegetation peaked 3 min after the initial male had left. Unfed and gravid females exited approximately 6 min after the first males. Conclusions Furvela tent-traps are suitable for the collection of outdoor biting A. arabiensis in Eritrea whilst the Prokopack sampler is the method of choice for the collection of resting insects. Constructing well-lit, rather than dark, animal shelters, may encourage otherwise endophilic mosquitoes to leave and so reduce their survival and hence their vectorial capacity.