The design of an algorithm for a programmable variable-rate counter for generating precise binary logarithmic functions is presented. The error in log2(l + x), as defined by Iog2(l + x) − x, may be considered as a set of straight lines whose slopes, either positive or negative, are chosen to be integral multiples of a binary fraction. By using a programmable counter whose rate is proportional to the slope of the line segments, the error is corrected. The circuitry is simple because no add operation is needed. The precision of the answer depends upon the number of bits used. In addition, an algorithm to synthesise the variable-rate up/down counter (VRU/DC), thus reducing the number of calculations, is given. It pinpoints the break points for this design, and also specifies the range covered by the segment for optimal precision. The algorithm can also be used to generate the antilogarithm.