This paper hypothesises that the growth in trade union membership has been considerably influenced by the economic conditions of workers and the activities undertaken by the unions to protect or improve these conditions. While the former may be represented by such variables as the growth in employment, prices and wages, the latter could be indicated by industrial action such as strikes called by the unions. A statistical model was developed and tested for the period 1930-69. This time-series analysis indicated that employment, wages, consumer price index and mandays lost contributed to growth of unionization. A cross-section analysis based on data for inter-state and inter-industry dispersion in union membership indicated that such dispersion was explainable by employment and mandays lost. There was also some influence of variation in wage-productivity ratio on the dispersion of trade union membership. It was concluded that the unions may make more intensive efforts to organise the workers in low unionised states and industries which may help in re-establishing the wage-productivity parity in these industries and states.