This series of observational surveys, to determine child safety seat use in Virginia, began in 1993 at the request of DMV officials. During all 4 years (there was no survey in 1995), data in metropolitan areas were collected at the same locations, at the same time of day and day of week, and according to the same criteria for determining use. In 1997, data collection was added for safety restraint use by occupants 4 to 16 years of age at the request of officials of the Department of Health, which is now responsible for the state's child safety seat program. In addition, data were collected on whether any booster seats in use were being used properly. Seven sites were also added in communities with a population between 50,000 and 100,000 (mid-size cities). For the entire vehicle, the 1997 metropolitan area child safety seat correct use rate was 54.1 %, incorrect use was 17.4%, and non-use was 28.5%. Non-use was greater in the front seats (42.1 %) than in the rear seats (25.2%). The western area had the highest non-use rate (50.0%) and the lowest correct use rate (32.1 %). Non-use and correct use in Northern Virginia, Central Virginia, and Tidewater were similar, with correct use ranging from 53.1% to 58.7% and non-use ranging from 24.8% to 29.1 %. Child safety seat use in the three areas categorized as mid-size cities (Charlottesville, Danville, and Lynchburg) was lower than in the metropolitan areas: correct use was 43.2%, incorrect use was 14.8%, and non-use was 42.0%, with Danville having the highest non-use rate at 61.9%. Non-use was higher in the front seats (66.7%) than in the rear seats (37.7%). Safety restraint/seat belt use by occupants 4 to 16 years old riding in the rear seats was very low. In the metropolitan areas, correct use was 34.9%, incorrect use was 2.3%, and non-use was 62.9%, with the western area having the highest non-use rate (69.3%). In the mid-size cities, correct use was 26.2%, incorrect use was 1.5%, and non-use was 72.3%, rates considerably worse than in the metropolitan areas, with Danville having a non-use rate of 84.7%. There was a high correct use rate for booster seats: 83.1% for the entire vehicle, 84.0% for the rear seats, and 77.8% for the front seats. The recommendations include the initiation of research to determine why child safety seat use is so low, a public information and education effort geared specifically toward child safety seat use, a special education and enforcement effort aimed at occupants 4 to 16 years of age, and frequent and continuous education and enforcement efforts because of changes in the population of the targeted groups.