Study data were obtained from the Puerto Rican Adolescent Survey, a two-wave panel data set composed of a sample of Puerto Rican male adolescents residing in the South Bronx of New York City. Sample screening yielded 1,170 eligible males, 1,077 (92 percent) of whom agreed to participate in the study. Respondent loss in the second wave, 1 year after the first wave, was 17 percent, resulting in an overall response rate of 76 percent. Confidential interviews were conducted in the respondent's home or elsewhere when home privacy could not be guaranteed. Delinquency measures included 27 offenses representative of the full range of offenses included in the Uniform Crime Report. Study findings indicated that peer involvement in delinquency was an important factor influencing youth involvement in delinquent behavior. The lack of a relationship between peer dimensions of intimacy and closeness and the number of friends who were involved in delinquent acts during the past year suggested that involvement with delinquent peers did not represent or refer to a specific, close type of experience between individuals. No significant relationship was found as to whether youths felt they were close to their friends, that friends cared, or that youths did not fit in with their peers. In youth involvement with delinquent peers, however, the nature of the exchange involved loneliness and estrangement, as friends were not willing to listen to problems. Time association appeared to be critical to understanding the dimension of involvement with delinquent peers. The lack of strength of intimacy and closeness in delinquent peer relationships suggested weaknesses in service programs that focus on shifting the goals of delinquent peer groups to prosocial activities. Other research on peer relationships and adolescent problem behavior is reviewed, and implications of the current study findings for delinquency etiology and prevention are discussed. 25 references and 2 tables