The paper examines Engestrom's version of activity theory. It seeks to locate this within the socio-economic and theoretical context in which notions of co-configuration and knotworking are set. Although this theoretical approach offers radical possibilities it is limited by its neglect of the wider social context in which activity systems are located as well as by its failure to address issues of power and social antagonism. The recovery of these neglected areas has been recognized by those who have critically engaged with activity theory. Such an approach would serve to re-centre the Marxist interest in social transformation and would minimize the tendency towards transformism integrally embedded in this particular version of activity theory. (Contains 2 figures and 2 notes.)