: Our tires are the only points of contact our vehicle will have to the road, therefore keeping them in prime condition will: maintain our vehicle’s handling capabilities, grip the road surface better in poor weather conditions, maximize tire life. Improve overall fuel economy. Although owner of vehicle can check air pressure and tread depth, but the most are unknown of the risks of they faced abrasion treadwear, cracks due to cuts or aging. Therefore, it’s important to be able to identify and fix signs of uneven wear and damage before they compromise your vehicle’s safety. We have used image of tire exterior (thorns, sidewalls...) identification car bad tires model, with image recognition technique. The proposed algorithm includes image normalization, deep learning convolutional neural networks, image classification with common errors such as abrasion treadwear, cracks due to cuts or aging. There are not many works using CNNs on car tire condition assessment which mostly used pretrained models and existing CNN architectures. In this paper, a CNN architecture with 5 layers is proposed and used to car tire condition classification task. We use of Exponential Linear Unit (ELU) and Rectified Linear Unit (ReLU) as the non-linearity function of CNN is experimented for comparison. The model is applied on image of tire exterior datasets and the the results is effective for image of tire exterior classification with accuracy impressively. We raise awareness operator of the risk of abrasion treadwear, cracks due to cuts or aging and alert vehicle safety by themselves.