Magnetic resonance (MR) imaging confronts the user with a large number of options that can critically affect image quality. Among the most important parameters over which the user may exercise control are receiver frequency, tip angles, imaging and compensatory gradient strengths, the number of signals averaged, the number of phase encoding gradient steps, and the repetition times. All of these parameters were systematically varied and optimal values determined for a 0.5T [tesla] Teslacon system, a task which employed both phantoms and normal volunteers. Under the normal conditions of tuning, the z-compensatory gradient presents the most critical adjustment. The compromise between image quality and time of study was also carefully examined, with the following determined as best: 192 gradient steps plus 2 signal averages for the head coil, or 192 gradient steps plus 4 signal averages for the body coil.