Data were collected from approximately 2,700 case investigations drawn randomly from police and laboratory films in four jurisdictions: Peoria, Ill.; Chicago, Ill.; Kansas City, Mo.; and Oakland, Calif. The process which controls the recognition, collection, and examination of forensic evidence is addressed, and incident variables which affect the gathering of physical evidence are considered, including the time lapse between the discovery of the crime and its report to police, the extent of physical interaction between the offender and the scene or the victim, and the presence of witnesses. Blood, hair, firearms, and fingerprints are identified as the forms of physical evidence most frequently collected and examined. Findings indicate that clearance rates for robberies and burglaries are significantly higher in investigations where physical evidence is examined. Additionally, forensic evidence has its greatest impact on cases which traditionally have the lowest solution rate -- cases with suspects neither in custody nor identified at the beginning of the investigation. The effects of scientific evidence on the clearance and prosecution of aggravated assaults are found to be less pronounced, and in many cases, not significantly different from cases with no forensic evidence. A number of recommendations, aimed principally at the pistol, detective, and crime scene and crime laboratory functions, are presented. These recommendations and suggestions for future research are intended to focus limited police and scientific resources on those investigations where physical evidence can have the greatest impact. Also included are 53 tables, 9 figures, and 46 references.