This study was conducted on identifying the “Dynamics of Voting Behaviour in Pakistan: A Case Study of Selected Districts in Khyber Pakhtunkhwa (2002- 2013)”. It is a quantitative study aimed at knowing the dynamics of voting behaviour in the past three elections (2002, 2008 and 2013), with major purpose to see the changing trend in the mentioned elections. It was proceeded with 292 respondents who had cast vote in any of the target elections and they were asked through Likert scale questionnaire under purposive sampling technique. Originally, the sample size was 300 but none of women voted in sampled election in District Dir Upper due to conservative culture and consensus of all political parties not to let women cast their vote which reduced the sample size from 300 to 292. Major objectives of the study were to identify the association between social factors and voting behaviour in Khyber Pakhtunkhwa; to see the relationship of political factors with voting behaviour; and to explore the association between propaganda factors and voting behaviour in Khyber Pakhtunkhwa. The study includes voting behaviour as dependent variable whereas independent variables include social factors, political factors and propaganda factors. Univariate and bivariate analyses were drawn. The study found a significant association of social factors such as family, biradari, feudalism, religion, political discussion at home among family members and personal interests with voting behaviour. However, ethnicity, social class and religious sects were not found in association with voting behaviour. The changing scenario was observed as the percentage of vote casting as personal decision increased with every successive election. Among social factors, family was identified as a major influencing factor of voting behaviour but its importance got diminished with each passing election and similarly the influence of biradari influence remained restricted to some areas. Religion was highly significant in election 2002 but lost its ground to other factors in the subsequent elections. The influence of landlordism was although slight but consistent in the sampled elections. The study also reveals that political factors such as party affiliation, performance evaluation, issues evaluation and party ideology were significantly associated with voting behaviour. Voting based on candidate affiliation was dropped while voting based on party affiliation increased in each successive election. Party leadership was non-significant in election 2002 and 2008 but came up as a significant factor in election 2013. Rest of the political factors remained consistent in influence in the sampled elections. Propaganda factors including public speeches by the politicians, news items in newspapers, TV talk shows and campaign on social media were significantly associated with the voting behaviour mainly in election 2013. However, door to door canvassing decreased in election 2013. Campaign on social media influenced none of the respondents in the elections 2002 and 2008, while it remained highly significant in election 2013. In addition, the study finds least women’s participation in the voting process. The study recommends that the local political leadership as well as Election Commission of Pakistan should come forward to actively launch an awarenesscreating movement regarding the importance of vote in the sampled areas towards increasing the participation ratio of women’s voting.