A vehicular ad hoc network is an emerging technology based on MANET which offers various services for safety and security of people on road. Vanet is helpful in providing secure and intelligent transportation services. But asvanet is dynamic in nature and topology of vehicles changes very rapidly it is also highly vulnerable to security attacks as compared to mobile ad hoc networks. One of the most threatening security attacks in vanet is Denial of Service (DoS) attack and Sybil attack. In DoS attack the attacker prohibit legitimate vehicle to use the network services by sending huge number of fake message so that network gets busy and it becomes unavailable for legitimate vehicles and in Sybil attack the attacker sends lots of fake messages in vanet to misguide other vehicles by changing its id. In this research paper we have used a two stage approach to detectthe DoS and Sybil attack in VANET. In first stage we have calculated the threshold value of number of vehicles and in second stage detected the attack by categorizing the vehicles based on number of hope count and determining and comparing the packet delivery ratio.