Although Web-based instruction offers an advantageous approach to medical education, few studies have addressed the use of Web-based education to teach clinical content at the postgraduate level. Even fewer studies have addressed clinical outcomes after the Web-based instruction, yet postgraduate training requirements now focus on outcomes of training. A randomized trial was conducted to compare knowledge of postgraduate year (PGY) 1 residents after Web-based with that after paper-based instruction and to compare residents' clinical application of their instruction using unannounced standardized patients (SPs) and unannounced activated standardized patients (ASPs). PGY 1 residents were assigned to a month-long ambulatory rotation during which they were randomized as a block to Web- or paper-based instruction covering the same four geriatric syndromes (dementia, depression, falls, and urinary incontinence). Outcome measures were mean change scores for before and after testing and scores from SP and ASP clinical encounter forms (checklist, chart abstraction, and electronic order entry). Residents who completed the Web-based instruction showed significantly greater improvement on the knowledge tests than those who received paper-based instruction. There were no significant differences in the scores from the SP and ASP clinical encounters except that the chart abstraction score was better for Web-based group than the paper-based group for dementia. Web-based instruction is an educational tool that medical residents readily accept and can be used to improve knowledge of core geriatrics content as measured using immediate posttesting. More-intensive educational interventions are needed to improve clinical performance by trainees in the care of older patients.