There are two main purposes of the development of Magnetic Resonance Imaging (MRI) techniques. One is to increase scanning speed; another is to increase image resolution. An important factor that affects the limitation of acquisition time and resolution is the signal-to-noise ratio (SNR). There are two main sources of MRI noises: sample and receiving coil. Thus one way to reduce the coil noise is using non-resistive high-temperature superconducting (HTS) coils. Previous works on HTS coils can be put into two categories: tape coils and thinfilm coils. Tape coils are suitable for surface coils with diameters larger than 6 cm or volume coils; thin-film coils are more suitable for a smaller planar surface coil. In this work we began from making tape coils to be used on Bruker 3T MRI system, to the design of thin-film surface coil pattern and the cryostat. Commercialized Bi2Sr2Ca2Cu3Ox tape was used to fabricate a HTS tape coil with a diameter of 7 cm. Tested by phantoms of different conductivity, we got a plot of HTS SNR gain over an equivalent copper coil that was in agreement with theoretical prediction. A SNR gain of about 2.4 can be obtained from the HTS tape coil over a conventional copper surface coil in kiwi fruit images and the acquisition time was expected to reduce to one-fifth the original time when keeping the same SNR. This is very beneficial when doing experiments that usually takes a long time (as conventional diffusion imaging); the saved time length can be up to several hours. YBa2Cu3O7 (YBCO) coated on LaAlO3 substrate was chosen to make the smallscale surface coil. Successful simulations of double-sided YBCO coils on a 2cm*2cm LaAlO3 substrate were done by Sonnet 6.0a. A spiral pattern of the HTS film that resonates at 125.3 MHz was decided, and its B1 field was also found out. We also designed a G10 glass fiber cryostat to be fit in the Bruker S116 mini gradient system. This cryostat can keep the HTS coil at 84 K for about 52 minutes. The SNR gain of HTS coils increases as the sample size goes down, so we can expect it to improve the SNR significantly when implemented in molecular imaging that needs very small FOV and high resolution. We also aim on developing multi-channel HTS phased arrays, together with other techniques that can raise signal voltage (for example, xenon imaging), to achieve the goal of ultra fast, high resolution and high SNR imaging.