This paper deals with a general framework for a posteriori error estim
ates in boundary element methods which is specified for three examples
, namely Symm's integral equation, an integral equation with a hypersi
ngular operator, and a boundary integral equation for a transmission p
roblem. Based on these estimates, an analog of Eriksson and Johnson's
adaptive finite element method is proposed for the h-version of the Ga
lerkin boundary element method for integral equations of the first kin
d. The efficiency of the approach is shown by numerical experiments wh
ich yield almost optimal convergence rates even in the presence of sin
gularities.