The present study makes an extensive use of secondary data on price of major crops in Bangladesh over the period of 20 years. To understand developments in marketing system, a 20 year period is considered to be long enough. The study covers the time period of 20 years from 1986-2005 (as the latest data available). 1986-1989 period was before full implementation of privatization policies. 1990 - 2005 period was after implementation of privatization policies. The later period fell in the structural adjustment reform period. Privatization of irrigation equipment and fertilizer distribution was introduced during this period. Besides, subsidies on fertilizer and other inputs were also largely withdrawn during this period. Further, output markets were liberalized and private sector was allowed to import food grains. That is, the economy essentially has thrived during this period based on functioning of markets. Selection of Prices In the present study harvest and lean period one month price of selected crops (average of four weeks) have been taken for analysis. Off-season (lean period) price indicates the wholesale price in processed form prior to one month of harvest. For correlation and cointegration analyses, deflated rather than nominal prices were used to assess the relationship in terms of real price changes. This is imperative when using time series data spreading over a 20 year period, when inflationary price rises were substantial. Agricultural raw materials prices (index) for the corresponding period were used as deflator. For a district market price, the respective divisional agricultural raw-materials prices index was used as the deflator, as the index is only available up to the divisional level. Choice of markets For Aman (HYV), seven leading district markets have been selected. These are Dhaka, Kishoregonj, Rangpur, Satkhira, Potuakhali, Moulvibazar and Feni. For Boro (HYV), Dhaka, Kishoregonj, Joypurhat, Jessore, Barisal, Comilla and Sunamgonj, and for wheat, Dhaka, Faridpur, Thakurgaon, Meherpur, Bhola, Chandpur and Hobigonj are selected. For each crop Dhaka is selected as the central market. For mustard, Dhaka, Myminsingh, Rajshahi, Khulna, Barisal, Hobigong and Chittagong are taken. For lentil Dhaka, Mymensingh, Rajshahi, Khulna, Chandpur, Hobigonj and Chittagong are selected. All these are leading growing areas, for that matter, largest markets for the respective cropsAnalytical Techniques Measures of the spatial integration of markets by correlation coefficient The degree of correlation between prices in various markets is taken as an index of the extent to which the two markets are integrated. A higher degree of the correlation co-efficient indicates a greater degree of integration, at least in terms of the pricing of the product between the local market and central market, vice versa. The correlation coefficient in the price of a commodity in any two markets is assumed unity under conditions of perfect spatial price integration. A correlation co-efficient of 0.81 or more is a high degree of inter market price relationship because, in such a case, 81 percent or more variation in the prices in one market is associated with that in another market, and that the remaining 19 percent variation may be assumed to stem from transportation, information lag and data bottleneck ( Lele, cited in Acharya & Agarwal, 1994, p. 217). In this study we assumed `r' value of 0.70 for two markets as highly correlated in price changes as we have used deflated prices rather than nominal prices of crops and followed test of significance. Revallian (1986), Ahmed and Bernard (1989) all used nominal prices in their analysis. The simple correlation coefficient for the prices in each pair of selected markets has been estimated. There is often problem of spurious correlation between times series variables-when there is effect of inflationary growth (if it is price variable) or variables show same growth trends or for some other reason. So, advanced method of assessing market integration (by cointegration test) was also used in this study.