It often appears that leading politicians in Belgium consider the results of the municipal elections in the light of national polities. They stick to the thesis that the municipal poll-results, at least in the bigger towns, are more and more influenced by the constellation of the nation's politics.Is it really so that the municipal elections indicate the hearings of the national political situation? Can one draw conclusions from the results of these elections as if they were national ones? And can one,any how, compare municipal to parliamentary elections? By comparison of the results of municipal elections 1964 and of parliamentary elections 1965 it was checked which shifts in party-choice havehappened during this short period of eight months. If no oscillations, or only a few, were detected between both elections, one could conclude that the municipal elections 1964 indicated indeed the hearings of the parliamentary elections 1965.Successively, the national and provincial results of these elections were compared and the party-shifts on national and provincial level were calculated.To make a relevant comparison between the election-results on a lower level, a comparable basic unity had to be found. As there were no municipal data available at parliamentary elections on one side, andonly municipal results at the municipal elections on the other side, the least possible unity for which parliamentary election-results are known, the electoral canton namely, was chosen as a comparison-basis. For that purpose however the results per party had to be additioned in all municipalities belonging to one canton.Part of the electoral cantons was unfit for use as comparison-material for two reasons: the highly varied and often strongly local-coloured political party-structure on one hand, the big number of municipalitieswithout elections on the other hand. So we were bound to make a choice out of 212 electoral cantons. Finally the cantons with a maximum of 4 municipalities were chosen, which limited the number of cantons to 28. These cantons were classified by degree of urbanisation according to the typology of W. Van Waelvelde and H. Van der Haegen.In that classification the percentages of votes in favour of the political parties, at the occasion of these elections, were tabulated and compared.The participating parties and lists were grouped as much as possible around the traditional parties to which they were most related. So we distinguish in Flanders: CVP, BSP, PVV, VU, CPB and other parties ;in the Walloon region: PSC, PSB, PLP PCB, French-speaking lists and other parties.For this analysis we also thought it was relevant to control separately the shifts of the electoral corps in the Flemish, the Walloon and the Brussels cantons. These shifts were then specified according to thedegree of urbanisation.To measure the size of the party-shifts for these elections, the external election-shift standard was calculated for the chosen Flemish, Walloon and Brussels cantons, with a special attention for the degree of urbanisation. Finally we examined which attitude was assumed by the government, the governmental and the opposition parties, in relation to the results of the municipal 1964 elections.It appeared that some notable party-shifts had been realized during the short period between the municipal elections of 1964 and the legislative elections of 1965. In general, a certain polarization has taken place due to a centrifugal vote-shift to the left and still more to the right. The direction of vote-shifts, which had shown at the municipal elections of 1964, was affirmed at the legislative elections of 1965 and for some parties, CVP and PVV namely, it was even accentuated. The parallelism between both elections in relation to the direction of the vote-shifts did not mean however that the size of these shifts was the same everywhere. The image of the shifts was different according to linguistic region and degree of urbanisation.The analysis of the urbanisation-degree showed that the level of oscillations grew higher as the urbanisation-degree grew lower. Seen per linguistic region, the largest shifts had taken place in the Walloon cantons. According to the calculations of the electoral shift standards during the period 1964-1965, the lowest oscillations were noted in the Flemish and Brussels cantons, i.e. the voting-behaviour of the big agglomerations during the municipal elections of 1964 were the closest to the national electoral pattern. Seen that way they were, up to a certain degree, a value-measure for the general policy.Finally the remark should be made that the limitation to the two above-mentioned elections does not allow any generalization of the obtained conclusions. The short period between those elections was itself an exceptional situation which may have been of influence on the results of the comparisons.