The viability of the Partial Credit Model (PCM) of G. N. Masters (1982) for scaling and banking individual essay prompts from a statewide writing assessment program was investigated. Preliminary analyses indicated the feasibility of using the PCM, the general polychotomous form of the Rasch model, with essay data from the Maryland Writing Test (MWT). Subjects were a cohort of ninth graders taking the MWT in the years 1984 through 1988. Data were the averaged narrative and explanatory scores from the administrations of the MWT in these years with their operative prompts and from the administration of the 1984 prompts used as anchor prompts each year in an equating study. Data were analyzed with the MSTEPS program of B. D. Wright et. al. (1988). Equipercentile and PCM banking procedures agreed on relative test difficulty in these data sets. Because of positive results from the analysis of link consistency among prompts and comparisons between equipercentile and PCM "equatings," the PCM appears viable for banking MWT prompts. Benefits of a prompt bank would include simplification of the prompt selection and evaluation process. Seven tables and two graphs present study data. (SLD)