The rate of hospitalizations due to ambulatory care-sensitive conditions (ACSCs) has been widely accepted as an indicator of access and quality of primary care. This study aimed to examine the trends and geographic variation of ACSC hospitalizations in US veterans health-care system, to identify factors associated with ACSC hospitalizations and to develop a quality indicator that can monitor access and effectiveness of primary care at hospital level. Using fiscal years 1997-2007 data, we found total ACSC hospitalizations per 1000 ACSC patients decreased by 58%; ACSC hospitalizations as percentage of total hospitalizations decreased 9%. However, significant geographic variations of ACSC hospitalizations remained and we found that adjustment of case-mix or confounding factors was essential in making meaningful comparisons among hospitals in a health-care system. Further, this study also reveals that low-income veterans still had higher ACSC hospitalization rates and patient travel time less than 30 minutes to the nearest VA providers was associated with fewer ACSC hospitalizations, which possess important policy implications.