The present study aims at describing the sampling plan, operational aspects and strategies used to optimize the field work of a cross-sectional, population-based study conducted in a southern capital of Brazil. For this purpose, the sample design, data collection instrument, selection of interviewers, pilot study, data collection, field logistics, quality control, consistency control, costs, and divulgation of results are herein described. The study's response rate was 85.3%. We found that the comparison of frequency measurements with and without self-assessment had no significant impact on the estimates, and that the design effect, estimated at 2, was sufficient for most calculations. The reproducibility of the questionnaire was satisfactory, with Kappa values and intraclass correlation coefficients ranging from 0.6 to 0.9. The strategies used to overcome operational problems, such as counting of households, use of maps, questionnaire structuring, rigorous organization of the field work and monitoring of the estimates were fundamental in conducting the study.