The behavior of aquatic stages of Aedes aegypti was evaluated in 200-liter metal drums to develop improved surveillance. When known numbers of larvae recovered with a rectangular net and with a standard dipper were compared, all stages except 1st instar were most numerous in the top 1/3 of the drum. Water temperature was the only significant variable affecting the vertical distribution of 4th-instar larvae; light intensity and pH were nonsignificant factors. Rust in the water was also thought to influence distribution towards the surface layer. When 100-400 immatures were released into a full drum, immatures were detected by sweeping and dipping the surface layer on 93.3 and 72.9% of trials, respectively. The correlation coefficients (r) of recoveries of known numbers of immatures by sweeping and dipping varied from 0.92 to 0.98 and from 0.38 to 0.89, respectively. The coefficients of variation were 2-5 times smaller for sweeping than for dipping. Recoveries from dipping were shown to be affected by the water volume in the drum. Counts of 4th-instar larvae from one sweep around the top are sufficient to assess productivity in drums.