In controlled/living radical copolymerization (atom transfer radical copolymerization in this study) and in any other living chain-growth copolymerization, the possible preferential addition of one of the comonomers onto the (macro) initiator-derived (macro) radical can affect the copolymer composition, especially at low conversion; this results in inaccurate comonomer reactivity ratio estimation by the classic approach. A new general approach is introduced in this article, which allowed us to exclude the influence of the possible preferential addition of one of the comonomers onto the (macro) initiator-derived (macro)radical on the copolymer composition at any conversion. According to this approach, copolymer chain grown during time t (t not equal 0) is considered to be, in fact, the macroinitiator terminated with one of the comonomers under study, which will further grow during the time interval Delta t' = t' - t [where any reaction time t' is considered to be grater than reaction time t, i.e. t' > t] from a comonomer mixture with composition of f(t) [where f(t) is the molar ratio of comonomer i to comonomer j in the comonomer mixture] at time t. In such a situation, it is possible to obtain individual comonomer conversions [x(i)(Delta t') and x(j)(Delta t')], the overall comonomer conversion [x(ov)(Delta t')], and the cumulative average copolymer composition for the copolymer formed during Delta t', from which more accurate comonomer reactivity ratios can be calculated by the various low- or high-conversion methods, depending on the overall comonomer conversion. (C) 2011 Wiley Periodicals, Inc. J Appl Polym Sci 122: 1341-1349, 2011