Given the dramatic increase in cell-phone only households and the difficulty of geographically-targeting cell-phone samples, address-based sampling (ABS) is the best available approach to obtain a representative sample within a small, geographically-defined area (e.g., neighborhood or community). When the data collection mode is telephone, a telephone number must first be obtained for every sampled unit. This study valuates the effectiveness of record matching, a mailed household information sheet, and a web response option for such a purpose. The California Health Interview Survey (CHIS) oversampled 14 communities as part of The California Endowment’s (TCE) Building Healthy Communities (BHC) program. Initially 46% of sampled addresses were matched to a telephone number through two sampling vendors. The remaining addresses were mailed a household information sheet asking for phone numbers associated with the address and a best time to reach someone. Addresses for which the originally-matched telephone number was not useful were also mailed household information sheets. This sheet also included a link to a website where sampled households could complete the form electronically, with more than 5 times as many returns coming via mail compared to web. Sampled addresses received up to three mailings (initial letter, postcard, reminder letter) and a subset of non-respondents were eligible for in-person follow-up to collect the telephone number. We compare the rate of return for each community each contact, and between originally matched and originally unmatched cases. We also compare the screener and extended interview completion rates by timing of return (first mailing, postcard, second mailing) and mode of return (hardcopy, telephone, web, in-person).