This study is two pronged. Firstly, it analyses the recorded seismic data in order to ascertain the nature of seismogenic sources. Secondly, it focuses on the development of an empirical attenuation relationship for the study area (northern part of Pakistan). The analysis of seismic data demonstrates that the area has substantial potential to generate frequently low to moderate seismicity originating from relatively shallow crustal depth. The magnitude (ML) of the major earthquakes included in this study ranged 5.0-7.0. The Kohat (ML= 6.0), Astor Valley (ML= 6.2) and Kaghan Valley (ML= 5.6) earthquakes collectively inflicted around 100 fatalities, left 250 people injured and made around 2000 people homeless. However, Fatehjang (ML= 5.3) and Mangla (ML= 5.0) earthquakes caused no human casualty. Focal mechanism solutions of the Kaghan Valley and Kohat earthquakes suggested that the earthquakes were predominantly due to thrust faults, striking NNW-SSE and E-W, dipping 510 and 200 towards NE and N respectively, with minor strike slip component. The solutions agreed well with the nature of seismogenic sources. Similarly, fault plane solution indicated that Mangla earthquake was also owing to thrust, striking E-W and gently dipping towards N. In contrary, focal mechanism solutions for Astor Valley and Fatehjang earthquakes demonstrated that these earthquakes were caused by normal thrust faults, striking NE-SW and NNW-SSE, dipping 620 and 510 towards NW and NE, respectively. The solutions did not match with the nature of the faults causing these earthquakes. However, more seismic data are required to find a plausible explanation of that mismatch. The intensity maps of these earthquakes indicated intensity (MMI) in the range of V-VIII. The Muzaffarabad earthquake of magnitude (ML) 7.0 was however extremely devastating. Death toll was around 75000, injured people were 100,000 and around four million were displaced. Post earthquake seismicity included 6330 earthquakes within first month, 296 of which had magnitude (ML) greater than 4.0. Aftershocks were distributed over a length of 120 Km and were mostly concentrated in the northern extent of the Kashmir Thrust. Focal mechanism solutions demonstrated that Muffarabad earthquake was predominantly due to thrust, striking NNW-SSE and steeply dipping NE, with minor strike slip component. The solutions coincided well with slip nature of Kashmir Thrust. The intensity map showed remarkably high intensity up to XI. An attenuation relationship was developed for the study area. A comparison of peak ground acceleration (PGA) values assimilated by using newly developed relationship and observed PGA values for earthquakes of 7.6, 6.4 and 5.0 magnitude showed a good agreement between the observed and assimilated PGA values. However, new relationship slightly overestimated the near field PGA values for earthquakes of magnitude 7.6. This discrepancy may be due to the possible site nonlinearity effects and dearth of near field high quality strong motion data, especially pertaining to earthquakes of greater magnitude. Likewise, a comparison of PGA values predicted by the new relationship and other relationships with observed PGA values for earthquakes of 7.6, 6.4 and 5.0 magnitude demonstrated better fit between the observed PGA values and values assimilated by the new relationship. This suggested the viability of newly developed attenuation relationship to predict peak ground acceleration for earthquakes of wider range magnitude and longer source to seismic station distances.