WHEN THE CENTER FOR URBAN AFFAIRS AT NORTHWESTERN UNIVERSITY GATHERED INFORMATION FOR TWO INVESTIGATIONS OF THE IMPACT OF CRIME ON THE LIVES OF CITY DWELLERS, FUNDS WERE LIMITED. IT WAS, THEREFORE, NOT POSSIBLE TO SEND PERSONAL INVESTIGATORS INTO EACH OF THE CITIES CHOSEN FOR DATA COLLECTION (SAN FRANCISCO, PHILADELPHIA, AND CHICAGO). INSTEAD, COMPUTER-GENERATED TELEPHONE NUMBERS BASED ON THE DEMOGRAPHIC CHARACTERISTICS OF EACH CENTRAL TELEPHONE OFFICE WERE USED. THIS TECHNIQUE IS DESCRIBED IN DETAIL, AS ARE THE TECHNIQUES THE TELEPHONE SURVEYORS USED IN GATHERING DATA. THE DEMOGRAPHIC DATA GATHERED BY PHONE AGREED CLOSELY WITH CENSUS DATA FOR THE AREA AS A WHOLE. THE ONLY DISCREPANCY WAS THE HIGHER PERCENTAGE OF HIGH SCHOOL GRADUATES FOUND IN THE PHONE SURVEY. ABOUT 6 PERCENT OF THE INTERVIEWS WERE CONDUCTED IN SPANISH. AN ADVANTAGE OF THE SYSTEM WAS THAT A SPANISH-SPEAKING INTERVIEWER COULD BE AVAILABLE TO TAKE OVER IF ENGLISH PROVED DIFFICULT FOR THE RESPONDENT. ABOUT 2.7 PERCENT OF THE 5,121 CONTACTS WERE JUDGED 'NOT COOPERATIVE,' ABOUT 2.9 WERE JUDGED 'INACCURATE,' AND 5 PERCENT REFUSED TO COOPERATE. THESE FIGURES ARE SIMILAR TO THOSE FOUND FOR PERSONAL INTERVIEWS. CHARTS EXPLAIN THE SURVEY TECHNIQUE AND TABLES GIVE RESPONSE RESULTS. IT IS CONCLUDED THAT RANDOM NUMBER DIALING IS AN EXCELLENT SURVEY TOOL WHEN FUNDS ARE LIMITED. APPENDIXES DESCRIBE THE TECHNIQUES USED TO CODE OPEN-ENDED ITEMS, GIVE NOTES ON ANALYSIS FILES, AND REPRODUCE THE SURVEY QUESTIONNAIRE. (GLR)