Multimodal Large Language Models (MLLMs) such as GPT-4V and Gemini Pro face challenges in achieving human-level perception in Visual Question Answering (VQA), particularly in object-oriented perception tasks, which demand a fine-grained understanding of object identities, locations, or attributes, as indicated by empirical findings. This is mainly due to their limited capability to effectively integrate complex visual cues with textual information and potential object hallucinations. In this paper, we present a novel approach, Joint Visual and Text Prompting (VTPrompt), that employs fine-grained visual information to enhance the capability of MLLMs in VQA, especially for object-oriented perception. VTPrompt merges visual and text prompts to extract key concepts from textual questions and employs a detection model to highlight relevant objects as visual prompts in images. The images, processed with visual prompts, alongside text prompts are subsequently fed into MLLMs to produce more accurate answers. Our experiments with GPT-4V and Gemini Pro on three benchmarks, i.e., MME, MMB, and POPE, demonstrate significant improvements. Particularly, our method led to a score improvement of up to 183.5 for GPT-4V on MME and enhanced MMB performance by 8.17% for GPT-4V and 15.69% for Gemini Pro.