An instrument has been constructed for applying minute drops of liquid insecticide to selected parts of individual insects. The principle of the instrument, which is not original, is to express minute amounts of liquid from a hypodermic syringe, and blow the drop from the needle tip by a puff of air. This model is considered to have advantages over previous models working on the same principle. In particular, a very convenient form of fitting was developed for blowing the drop from the needle tip, and fittings of this type could be incorporated in existing similar applicators with little modification of the latter. The high directional accuracy of dropfire necessary for treating selected parts of small insects was achieved.An immersion method was developed for determining the sizes of individual drops of Shell oil P31delivered, and hence the uncontrolled variability of drop‐volume. It showed that the standard deviation in the volumes of drops intended to be equal in size was about 9% of a mean volume of 0.017 μ1. falling progressively to about 2% as the mean volume increased to 0.33 μl. These standard deviations, however, are of limited value, because the sizes of successive drops delivered appeared frequently to be correlated. A method in which the drops are delivered on to an oleophobic surface for measurement of the diameters of the oil lenses formed was found insufficiently accurate for determining the relative variability of drops of P31.A suction method was devised for manipulating, without anaesthetization or chilling, small beetles such asCalandraandTribolium.This enabled large numbers to be individually treated with the applicator in a reasonable time. A similar but slower method was used successfully with the larvae ofEphestia.The biological results achieved have been very satisfactory.The effects of uncontrolled variation in drop size on dose‐response data are disc